Viruses as living processes

ORE Open Research Exeter TITLE Viruses as living processes AUTHORS Dupré, John; Guttinger, S JOURNAL Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences DEPOSITED IN ORE 29 February 2016 This version available at http://hdl.handle.net/10871/20233 COPYRIGHT AND REUSE Open Research Exeter makes this work available in accordance with publisher policies. A NOTE ON VERSIONS The version presented here may differ from the published version. If citing, you are advised to consult the published version for pagination, volume/issue and date of publication Viruses  as  Living  Processes   John  Dupré1,*  and  Stephan  Guttinger1   1Egenis,  The  Centre  for  the  Study  of  Life  Sciences,  University  of  Exeter,  Byrne  House,  St  German’s   Road,  Exeter,  EX4  4PJ,  UK   *  Corresponding  author.  E-­‐mail:  [email protected]       Abstract   The  view  that  life  is  composed  of  distinct  entities  with  well-­‐defined  boundaries   has  been  undermined  in  recent  years  by  the  realisation  of  the  near  omnipresence   of  symbiosis.  What  had  seemed  to  be  intrinsically  stable  entities  have  turned  out   to  be  systems  stabilised  only  by  the  interactions  between  a  complex  set  of   underlying  processes  (Dupré,  2012).  This  has  not  only  presented  severe   problems  for  our  traditional  understanding  of  biological  individuality  but  has   also  led  some  to  claim  that  we  need  to  switch  to  a  process  ontology  to  be  able   adequately  to  understand  biological  systems.  A  large  group  of  biological  entities,   however,  has  been  excluded  from  these  discussions,  namely  viruses.  Viruses  are   usually  portrayed  as  stable  and  distinct  individuals  that  do  not  fit  the  more   integrated  and  collaborative  picture  of  nature  implied  by  symbiosis.  In  this  paper   we  will  contest  this  view.  We  will  first  discuss  recent  findings  in  virology  that   show  that  viruses  can  be  ‘nice’  and  collaborate  with  their  hosts,  meaning  that   they  form  part  of  integrated  biological  systems  and  processes.  We  further  offer   various  reasons  why  viruses  should  be  seen  as  processes  rather  than  things,  or   substances.  Based  on  these  two  claims  we  will  argue  that,  far  from  serving  as  a   counterexample  to  it,  viruses  actually  enable  a  deeper  understanding  of  the     1   fundamentally  interconnected  and  collaborative  nature  of  nature.  We  conclude   with  some  reflections  on  the  debate  as  to  whether  viruses  should  be  seen  as   living,  and  argue  that  there  are  good  reasons  for  an  affirmative  answer  to  this   question.     Key  Words:  Virus,  Life  cycle,  Symbiosis,  Mutualism,  Process  philosophy,   Substance  ontology     1.  Introduction   It  is  still  often  assumed  that  life  is  composed  of  discrete,  genetically   homogeneous,  organisms,  either  single  cells  or  the  descendants  of  a  single   originating  cell  in  the  case  of  multicellular  organisms.  This  assumption  accords   well  with  the  orthodox  metaphysical  thesis  that  the  world  is  composed  of  things,   or  substances.  These  things  are  typically  thought  of  as  fairly  stable  entities,  and   as  bearers  of  properties.  Although  these  properties  can  change,  some  subset  of   them  must  persist  if  the  entity  itself  is  to  persist.  Things  are  thought  of  as  having   reasonably  clear  boundaries,  and  their  important  properties,  the  properties  that   determine  their  continued  existence,  as  being  intrinsic,  i.e.  as  being  grounded  on   features  that  lie  entirely  within  those  boundaries.       Realisation  of  the  near  omnipresence  of  symbiosis,  however,  is  one  factor   that  has  presented  severe  problems  for  this  background  position  (Dupré,  2012,   chs.  7,  11).  Widespread  symbiosis  threatens  the  clarity  of  boundaries  between   organisms,  and  even  the  uniqueness  of  these  boundaries.  This  paper  starts  from   a  position  articulated  in  Dupré  and  O’Malley  (2009):  the  typical  living  system   consists  of  interconnected  and  collaborating  segments  of  many  genetically     2   distinct  lineages.  Humans,  for  instance,  comprise,  as  well  as  the  lineage  of   ‘human’  cells  derived  from  an  original  zygote,  numerous  lineages  of  symbiotic   bacteria,  archaea,  and  fungi.  These  vary  in  the  extent  to  which  they  are   mutualistic,  commensalistic  or  parasitic;  often  the  same  organism  can  play   different  such  roles  at  different  times  (Méthot  &  Alizon,  2014).  The  boundaries  of   the  organism,  which  may  or  may  not  be  taken  to  include  some  or  all  of  these   symbionts,  may  be  to  some  extent  indeterminate.  The  realisation  of  the   integrated  nature  and  blurred  boundaries  of  organisms  has  led  to  claims  that   traditional  (substance-­‐based)  metaphysical  accounts  of  individuality  should  be   replaced  with  a  process  ontology,  as  the  only  ‘philosophy  of  organism’  that  can   make  sense  of  the  biological  phenomena  as  we  now  know  them  (see  for  instance   (Henning,  2013)).   Whilst  the  adoption  of  a  process  ontology  might  be  thought  of  merely  as   an  epistemological  strategy  our  claim  here  is  an  ontological  one:  biological   systems  are  processes.1  It  is  not  just  that  biological  things  are  complexly   interrelated  with  other  biological  things.  These  relations  are  necessary  for  the   persistence  of  the  biological  system.  Whereas  persistence  is  the  default  state  of  a   thing,  the  persistence,  or  stability,  of  a  process  requires  explanation;  it  is  actively   maintained.  The  stabilisation  of  multicellular  organisms,  in  particular,  has  been   found  to  depend  not  only  on  internal  processes,  but  also  on  the  interactions   between  its  symbiotic  constituents,  which  leads  us  to  argue  that  all  or  most  of   these  should  be  seen  as  parts  of  the  overall  process  that  constitutes  the  organism.                                                                                                                   1  For  more  on  the  distinction  between  epistemological  and  ontological   processism  see  (Rescher,  1996).     3   The  organism,  thus  broadly  construed,  can  then  be  seen  as  a  stable  eddy  in  the   flow  of  interconnected  biological  processes  (see  also  (Dupré,  2012,  chs.  4,5)).   The  aim  of  this  paper  is  to  explore  the  role  of  viruses  in  relation  to  this   general  processual  view  of  life.  Viruses  have  usually  been  seen  as  distinct   individuals  that  are  entirely  competitive  among  themselves,  and  entirely  harmful   to  anything  else  unlucky  enough  to  be  affected  by  them.  Given  this   understanding  it  is  not  clear  how  viruses  could  fit  into  the  more  integrated  and   interdependent  picture  of  life  that  we  have  just  sketched.  They  are  rather  seen  as   distinct  entities  that  follow  their  own  intrinsic  (and  pathogenic)  agenda.   We  want  to  challenge  this  view  on  two  counts:  first  we  will  claim  that   viruses  should  be  understood  very  much  in  the  same  way  as  other  lineages  in  the   flow  of  living  systems.  As  we  will  discuss  in  Sections  2  to  4,  recent  research  in   virology  shows  that  there  are  also  ‘nice’  viruses.  Often,  as  is  very  familiar,  the   intersection  of  viral  processes  with  organisms  is  destabilising  and  pathogenic.   But  viruses  also  make  important  contributions  to  the  stability,  or  health,  of  the   hosts  they  intersect  with.  Symbiotic  systems  therefore  may  include  viruses  as   well  as  plants,  animals  and  microbes  (this  point  is  elaborated  by  Pradeu  (this   issue),  a  paper  highly  complementary  to  ours).      Second  we  will  argue  that  viruses  have  to  be  seen  as  processes.  Viruses   pass  through  an  intricate  and  specific  sequence  of  states  or  activities  that  must   be  seen  as  an  ongoing  and  repeated  series  of  cycles  (Sections  5  to  8).  Specific   stages  of  the  cycle  might  have  significant  stability  (for  instance  the  virion  stage),   but  this  stability  is  temporary,  and  the  fact  that  there  are  (perhaps  very  many)   such  temporarily  stable  entities  can  only  be  understood  by  reference  to  their     4   role  in  the  larger  process  that  is  the  virus.  This  processual  nature  of  viruses  will   be  elaborated  in  more  detail  in  the  second  half  of  the  paper.   Bringing  both  the  processual  nature  of  viruses  and  their  intermittent   ‘niceness’  to  the  fore  will  show  that  viruses  are  not  counterexamples  to  the   integrated  and  dynamic  picture  of  biological  systems  advocated  here  and   elsewhere  (Dupré,  2012).  Indeed,  the  example  of  viruses  serves  to  reinforce  (and   further  inform)  a  processual  view  of  biological  systems.  Viruses,  or  so  we  will   claim,  are  vital  and  omnipresent  constituents  of  the  larger  flow  of  interconnected   processes  that  make  up  biological  systems.       2.  The  Microbiome  and  its  Benefits   Not  long  ago,  it  was  standard  to  think  of  a  multicellular  organism  as  a  lineage  of   differentiated  cells,  originating  from  a  founder  cell,  typically  a  fertilised  egg.   Microbes,  especially  bacteria,  were  generally  thought  of  as  potential  enemies,   poised  to  invade  and  attack  the  multicellular  system.  It  gradually  became  clear,   however,  that  multicellular  organisms  are  typically  populated  by  vast  numbers   of  microbial  residents  and  that  these  often  do  little  harm.  Perhaps  they  are  just   passengers,  taking  advantage  of  a  warm  and  well-­‐resourced  niche.  But  it  was   also  clear  that  in  exploiting  these  resources  some  bacteria  also  provide  some   benefit.  In  the  case  of  animals  like  cows,  which  rely  on  digesting  such  recalcitrant   molecules  as  cellulose,  it  was  long  known  that  this  was  only  possible  with  the   help  of  resident  bacteria,  and  here  can  be  seen  the  beginnings  of  a  shift  in   perception  of  microbes  from  dangerous  threat  to  necessary  symbiont.       More  recently,  it  has  become  clear  that  microbial  symbionts  do  far  more   than  just  these  often  essential  contributions  to  digestion.  They  are  involved  in     5   the  modulation  of  development,  and  play  a  central  role  in  the  development  and   homeostasis  of  the  immune  system  (Round  &  Mazmanian,  2009;  Chu  &   Mazmanian,  2013;  Spasova  &  Surh,  2014).  They  have  even  been  found  to  connect   to  the  central  nervous  system  (Bravo  et  al.,  2012).  In  plants,  hugely  complex   systems  of  bacteria  and  fungi  modulate  the  interface  between  the  plant’s  roots   and  the  surrounding  soil  (Philippot  et  al.,  2013).  These  insights  have  contributed   to  a  major  philosophical  reconsideration  of  the  concept  of  the  biological   individual,  with  some  researchers  arguing  that  multicellular  organisms  are   typically  massively  symbiotic  individuals  or,  as  they  are  sometimes  known,   holobionts  (the  concept  of  holobiont  is  discussed  in  Mindell,  1992;  Rohwer  et  al.,   2002;  Rosenberg  et  al.,  2007).2  The  human  microbiome,  according  to  some,   consists  not  of  passengers,  but  of  parts  of  an  integrated  individual.       Importantly,  according  to  this  integrated  view  of  the  biological  individual,   the  organism  itself  in  its  stable  state  turns  out  to  be  a  product  of  a  myriad  of   interactions  between  host  and  microbes.  The  body  then  is  not  just  the  passive   and  pre-­‐existing  vessel  that  can  host  a  bacterium;  it  is  shaped  and  maintained  by   the  interaction  with  its  ‘guests’.     The  human  body,  however,  is  not  only  populated  by  bacteria,  archaea  and   fungi  but  also  by  viruses.  It  is  difficult  to  provide  a  good  estimate  of  the  number   of  virus  particles  within  the  human  body,  but  as  techniques  have  developed  for   finding  them,  results  have  been  more  or  less  consistent  with  the  analogical   inference  from  simpler  systems  studied  that  there  are  about  ten  times  as  many   virus  particles  as  cells  (Brüssow  &  Hendrix,  2002).  This  might  immediately  raise                                                                                                                   2  More  general  philosophical  discussion  is  provided  in  (Dupré  &  O’Malley,  2009;   Bouchard  &  Huneman,  2013;  Pradeu  &  Carosella  2006).     6   a  question  how,  if  viruses  are  as  uniformly  nasty  as  the  standard  view  supposes,   we  manage  to  stay  alive  at  all.     3.  Viral  collaborators?   As  pointed  out  at  the  beginning  of  this  paper,  there  is  evidence  that  the  resident   community  of  viruses  provides  services  to  biological  systems;  possibly  even  such   vital  services  that  we  should  consider  them,  like  many  bacteria,  to  be  integral   parts  of  complex  symbiotic  biological  organisms.     Apart  from  the  very  obvious  fact  that  they  frequently  fail  to  kill  us,  there   is  a  general  reason  for  supposing  that  the  vast  numbers  of  viruses  or  virus-­‐like   particles  found  in  the  human  body  are  an  integral  part  of  the  system  rather  than   a  reservoir  of  predators,  generally  kept  sufficiently  under  control  to  allow  the   system  to  function.  This  is  that  the  composition  and  size  of  the  virome  seem  to  be   remarkably  stable  (see  Section  5  for  a  more  detailed  discussion  of  the  term   ‘virome’).  If  viruses  were  primarily  hostile,  then  we  would  expect  their  numbers   to  oscillate  in  the  way  analysed  in  the  classic  Lotka/Volterra  predator/prey   models;  but  this  appears  not  to  be  the  case  (Reyes  et  al.,  2010;  Minot  et  al.,  2011;   Wylie  et  al.,  2012).  In  fact  the  virome  responds  gradually  and  in  a  systematic  way   to  changes  in  diet  (Minot  et  al.,  2011),  which  rather  suggests  a  positive  functional   response  to  environmental  change.     There  is  of  course  no  denying  that  the  way  viruses  reproduce  and   maintain  their  own  life  cycles  is  not  very  nice  for  the  host:  generally  when  they   have  used  a  cell  for  their  own  reproductive  purposes  they  kill  it  as  they  leave.   But  it  is  also  possible  that  this  very  killing  of  cells  is  functional  for  the  larger   system  of  which  a  virus  is  part.  Cells  after  all  die  all  the  time,  and  stabilising  a     7   complex  system  including  cells  of  many  kinds  will  very  centrally  involve   mechanisms  for  killing  cells  that  are  surplus  to  the  requirements  of  the  system.   Apoptosis,  or  programmed  cell  death,  is  for  this  reason  an  essential  process  in   multicellular  organisms.     The  fact  noted  above,  that  phage  populations  and  strains  remain  stable   within  an  individual  implies  a  stable  ecological  relation  between  these  and  their   bacterial  hosts,  and  it  seems  very  likely  that  this  is  functional  for  the  whole   system.  It  is  debated  whether  bacteria  have  any  equivalent  of  apoptosis,  and   since  bacterial  communities  almost  invariably  coexist  with  phages,  it  is  plausible   that  stable  relations  between  the  two  have  evolved  in  many  cases,  especially   where  these  are  functional  for  a  larger  system  of  which  both  are  part.   The  simplest  targeted  cell  destruction  role  to  attribute  to  viruses  would   just  be  the  killing  of  hostile  invaders.  And  indeed  there  is  evidence  for  such  a  role.   Barr  et  al.  (2013)  provide  evidence  for  a  coevolved  relationship  between   metazoan  mucosal  surfaces  and  phages.  They  describe  a  chemical  binding   between  elements  in  the  mucus  and  phage  capsids,  resulting  in  an  enriched  ratio   of  phages  to  bacteria  on  mucal  surfaces.  This,  they  plausibly  suggest,  protects  the   metazoan  from  external  bacterial  invasion  while  also  providing  the  phages  with   a  location  richly  provided  with  host  bacteria.  Importantly,  it  is  not  just  the   regulation  of  cell  numbers  that  matters  here.  The  bacteriophages  provide  a  non-­‐ host  derived  immunity  to  the  host  which  can  be  crucial  to  its  survival.  The   viruses  that  target  prokaryotes  indirectly  support  the  immune  system  of  the  host   by  taking  over  some  defence  work.     There  are  many  other  ways  in  which  viruses  are  now  known  to  help  their   hosts,  mostly,  it  seems,  by  playing  roles  similar  to  that  of  some  bacteria,  for     8   instance  in  the  regulation  of  gut  morphology  and  function  or  the  shaping  of  the   immune  system.  Such  ‘stabilising’  effects  of  viruses  on  their  hosts  are  discussed   in  more  detail  by  Pradeu  (this  issue)  and  we  will  not  further  dwell  on  this  point   here.       What  we  want  to  focus  on  in  the  next  section  is  another  manner  in  which   viruses  might  provide  benefits  to  their  host  or  to  populations  of  hosts  and  that  is   in  the  context  of  DNA  flow  between  cells.  In  these  examples  it  is  the  dynamic   nature  of  the  systems  involved  rather  than  the  stabilities  created  which  matters   most.     4.  The  Benefits  of  a  Flow  of  DNA   DNA,  we  often  think,  is  trapped  deep  in  the  heart  of  the  cell.  It,  or  its  copies,  is   passed  on  to  more  or  less  identical  cells  when  cells  divide.  In  this  process  the   DNA  doesn’t  escape  the  cell;  it  is  just  that  the  cell  splits,  and  the  DNA  moves  on  in   two  separate  prisons.     Surrounding  the  flow  of  DNA  is  a  penumbra  of  the  closely  similar  nucleic   acid  RNA,  the  molecule  through  which  DNA  sends  its  developmental  and   metabolic  orders  into  the  surrounding  biological  action  (indeed  some  of  it  has   long  been  known  as  “messenger”  RNA).  The  DNA,  however,  trapped  in  its   sequence  of  cell  nuclei,  is  separated  from  the  outside  world  by  two  membranes,   the  nuclear  membrane  and  the  cell  membrane,  precisely  regulating  the   molecular  milieu  in  which  it  lives.  In  the  process  of  mitosis,  or  cell  division,  the   DNA  briefly  escapes  its  inner  membrane  (at  least  in  those  organisms  undergoing   open  mitosis,  see  (Guttinger  et  al.,  2009)),  but  it  never  passes  beyond  the  outer   membrane  of  the  cell.  This  is  the  vertical  flow  of  nucleic  acids  through     9   evolutionary  time,  and  the  flow  often  conceived  as  accumulating  the  aeons  of   evolutionary  wisdom  and  carrying  it  forward  to  the  ever  more  complicated  and   well-­‐adapted  organisms  that  exist  today.   But  there  is  also  a  horizontal  flow.  Viruses  carry  on  their  life  cycles  by   entering  cells,  reproducing  there,  then  leaving  to  find  new  cells  to  invade.   Although  typically  the  viral  DNA  does  not  remain  in  the  invaded  cell,  which  is   very  often  killed  in  the  course  of  the  encounter,  sometimes  it  does.  Retroviruses,   for  instance,  have  a  single-­‐stranded  RNA  genome  which  is  reverse-­‐transcribed   after  infection  into  double-­‐stranded  DNA  and  inserted  into  the  host  genome.   After  insertion  the  viral  DNA  is  treated  by  the  host  like  its  own  DNA,  meaning  it   is  transcribed  and  reproduced  together  with  the  rest  of  the  host  genome.   Importantly,  in  some  cases  retroviruses  can  become  entrapped  in  the  host   genome  and  turn  into  what  is  known  as  endogenous  retroviruses  (ERVs).  Over   time  this  has  resulted  in  a  significant  addition  of  DNA  sequence  to  mammalian   genomes  and  it  is  estimated  that  up  to  8%  of  the  human  genome  actually  consists   of  ERVs  (Griffiths,  2001;  Lander  et  al.,  2001).  This  means  that  significant   proportions  of  the  DNA  in  eukaryotic  organisms  originally  entered  its  cell-­‐line  by   way  of  a  virus.     This  horizontal  flow  of  nucleic  acids  is  vast.  Think  of  the  1029  or  so   bacteria  that  are  estimated  to  live  in  the  Earth’s  oceans,  an  inconceivably  large   number,  but  not  as  large  as  the  1030  phages,  the  viruses  that  prey  on  them   (Whitman  et  al.,  1998;  Suttle,  2007).  1030  phages,  it  is  estimated,  would  cover   hundreds  of  light  years  placed  end  to  end,  so  the  scale  of  this  horizontal  flow  is   vast  indeed  (see  also  Morgan,  this  issue).  Not  much  of  this  would  need  to   establish  itself  in  the  vertical  flow  to  make  a  significant  impact  on  the     10   evolutionary  process.  And  in  fact  it  is  becoming  increasingly  clear  that  there  is   just  such  an  impact.  Among  microbes  (bacteria,  archaea,  and  to  some  extent   single-­‐celled  eukaryotes)  lateral  gene  transfer,  to  a  large  extent  mediated  by   viruses,  is  undoubtedly  important,  and  has  led  many  theorists  to  question   whether  the  traditional  representation  of  evolution  as  a  ‘tree  of  life’  portraying   the  vertical  relations  between  kinds  of  organisms  is  defensible  or  even   meaningful.  Microbes  have  acquired  their  DNA  from  a  variety  of  ancestors   (Bapteste  et  al.,  2009).     Horizontal  exchange  of  DNA  between  bacteria  can  also  be  mediated  by   plasmids,  circular  pieces  of  DNA  that  are  distinct  from  the  main  genome  and   which  are  typically  found  in  bacteria  (for  an  overview  see  (Sherrat,  1974)  or   (Smillie  et  al.,  2010)).  Like  viruses,  plasmids  can  move  from  cell  to  cell.  Some   have  the  ability  to  generate  a  pilus,  a  process  that  enables  them  to  transfer  to   another  cell  by  the  quasi-­‐sexual  process  of  conjugation.  Others  rely  on  the   assistance  of  other  plasmids.  Well-­‐studied  plasmids  provide  their  host  cells  with   useful  functions,  such  as  antibiotic  resistance,  virulence,  or  nitrogen  fixation.   They  thus  can  provide  a  reserve  of  access  to  such  functions  within  a  bacterial   community  without  the  energetic  costs  of  every  member  of  the  community   maintaining  these  resources.  The  ability  of  a  community  to  respond  to  a  threat   such  as  an  antibiotic  by  distributing  a  resource  held  by  a  few  members  of  the   community  is  far  more  efficient  than  requiring  every  member  of  the  community   to  be  independently  equipped  to  respond  to  any  likely  threat.  Indeed,  there  is  an   obvious  parallel  with  the  division  of  labour  between  cells  of  different  kinds  in  a   multicellular  organism,  a  parallel  that  lends  some  support  to  the  hypothesis  that   bacterial  communities  may  themselves  be  best  seen  as  multicellular  individuals.       11   This  mode  of  function  is  well  established  for  the  case  of  microbial   communities  such  as  biofilms.  But  as  we  have  already  noted,  complex  eukaryotes   are  all  or  almost  all  in  fact  symbiotic  systems  involving  microbial  consortia  as   well  as  a  lineage  of  eukaryotic  cells  (Dupré  &  O’Malley,  2009).  So  if  transfer  of   genetic  material  is  important  to  the  functioning  of  microbial  communities,  it  is   important  to  the  functioning  of  us.  When,  for  example,  we  consume  antibiotics  to   combat  harmful  bacteria  that  are  damaging  us,  antibiotic  resistance  plasmids   that  help  our  symbiotic  communities  resist  this  threat  are  likely  to  be  good  for   us—though  needless  to  say  they  also  carry  the  risk  of  being  co-­‐opted  by  the   pathogenic  bacteria  the  antibiotics  are  intended  to  destroy.     It  may  be,  in  fact,  that  the  virome  functions  as  a  vast  storehouse  of  genetic   resources.  Minot  et  al.  (2011)  compared  the  results  of  a  metagenomic  analysis  of   the  human  virome  with  a  database  of  known  antibiotic  resistance  genes,  and   found  614  matches.  Wylie  et  al.  (2012)  sum  up  the  situation:  “like  bacterial   plasmids,  bacteriophages  serve  as  reservoirs  for  mobile  genetic  elements  in   bacteria.  In  turn,  this  suggests  that  bacteriophages  may  affect  human  health  by   contributing  to  or  changing  the  metabolic  capabilities  of  the  resident  bacterial   community.”  As  one  of  us  has  suggested  elsewhere  the  best  way  to  think  of  the   human  genome  itself  may  be  as  a  database  or  library  of  resources  that  can  be   used  in  multiple  ways  by  the  cell  (Dupré,  2005;  see  also  Noble,  2006).  From  this   point  of  view  we  might  then  see  the  virome  as  a  reserve  warehouse  of  genetic   resources;  or  to  vary  the  metaphor  slightly,  the  genome  as  the  open  shelves  of   the  library,  with  the  virome  as  the  stacks—though  the  genome  here  must  be   interpreted  to  include  the  99%  of  the  human  genes  that  reside  in  the  microbiome.     12   An  even  more  speculative  thought  is  that  the  ability  of  microbes,   specifically  our  symbiotic  microbiome,  to  recruit  genetic  resources  from  the   biotic  environment  may  be  a  much  more  efficient  way  of  responding  to   environmental  contingencies  than  evolution  by  random  genetic  variation  and   selection.  One  minor  example  of  this  is  the  acquisition  of  genes  from  marine   bacteria  by  the  gut  bacteria  of  Japanese  people  that  enable  them  to  digest   seaweed  (Hehemann  et  al.,  2010).  The  diversity  of  roles  played  in  the  functioning   of  complex  symbiotic  systems  by  bacterial  consortia  suggests  that  this  could  turn   out  to  be  a  frequent  type  of  occurrence.  If  so,  it  is  likely  that  viruses  or  plasmids   are  the  agents  responsible  for  this  kind  of  traffic.   In  summary  we  see  then  that  in  addition  to  the  vertical  travel  of  DNA   down  the  generations  of  a  cell  there  is  a  vast  stream  of  nucleic  acids  moving   between  cells.  And  though  generally  the  interactions  between  these  streams  are   transient,  there  is  enough  long-­‐lasting  interaction  that  a  substantial  proportion   of  the  DNA  in  the  vertical  stream  arrived  there  by  intersection  with  the   horizontal  stream.  What  once  seemed  to  be  an  isolated  system  going  in  one   direction  only  (vertical  DNA  flow)  is  actually  much  more  open  and  flows  in   several  directions  at  the  same  time.     Given  this  tight  interconnection  between  viruses  and  their  hosts  it  seems   at  least  a  plausible  hypothesis  that  viruses  in  complex  multi-­‐organismic  systems   are  vital  functional  parts  of  the  whole.  As  we  have  seen  in  the  first  part  of  this   paper,  they  may  well  play  essential  roles  in  regulating  the  numbers  of  different   cell  kinds,  especially  microbial  kinds,  and  immunological  roles  in  eliminating   deleterious  cells,  certainly  potentially  hostile  microbes.  They  may  well  also  play   important  roles  in  mediating  the  transfer  of  genetic  resources,  surely  between     13   symbiotic  microbes,  and  possibly  even  between  the  latter  and  cells  in  the   eukaryotic  host.  And  more  generally,  as  discussed  by  Pradeu  (this  issue),  viruses   can  also  support  the  development  of  their  hosts  and  help  them  survive  under   challenging  conditions.     5.  The  Human  Virome   In  the  previous  sections  we  have  repeatedly  referred  to  the  ‘virome’  without   specifying  in  detail  what  is  meant  by  this  term.  In  part  this  is  because  the   question  of  how  to  define  the  virome  has  no  straightforward  answer.  The   problems  with  the  term,  as  we  want  to  show  in  this  section,  hang  together  with   the  issue  of  how  we  think  of  the  ontological  status  of  viruses  (which  we  will   discuss  in  Section  6).     The  suffix  ‘–ome’  is  now  widely  used  to  refer  to  a  class  of  biological   entities  of  some  general  kind  found  in  a  particular  biological  context.  The  ‘liver   transcriptome’  for  instance  is  the  set  of  transcripts  (i.e.  the  RNAs  corresponding   to  transcribed  DNA  sequences)  that  can  potentially  be  found  in  liver  cells.   Following  this  use  of  the  suffix  the  set  of  viruses  found  associated  with  humans   would  therefore  be  called  the  ‘human  virome’.     We  indeed  find  the  term  used  in  this  sense  in  the  life  sciences,  for  instance   in  Delwart  (2013),  who  writes:  “It  is  now  conceivable  that  all  viral  species   commonly  infecting  human[s]  (i.e.,  the  human  virome)  will  soon  be  determined”.   The  fact  of  infection  itself  does  not  tell  us  whether  the  presence  of  the  virus  is   detrimental,  beneficial  or  of  no  further  consequence  to  the  host.  It  also  does  not   tell  us  whether  the  infection  results  in  a  prolonged  or  only  temporary  presence   of  the  virus  in  the  host  system.  (The  only  restriction  in  this  definition  lies  in  the     14   term  ‘commonly’,  which  suggests  that  ‘uncommon’  human  viruses  will  not  be   counted  as  part  of  the  human  virome.)   Other  scientists  use  the  term  virome  in  a  more  restricted  sense,  referring   only  to  the  viral  component  of  the  microbiome,  the  latter  being  defined  as  the  set   of  all  microbes  (bacteria,  archaea,  fungi  and  viruses)  stably  associated  with  an   organism  (Wylie  et  al.,  2012;  see  also  Lecuit  &  Eloit,  2013).   Defining  the  virome  as  a  sub-­‐set  of  the  microbiome  implies  a  different   understanding  of  the  set  of  viruses  to  which  it  refers.  The  human  microbiome,  as   we  have  noted,  is  not  just  a  list  of  microbes  that  infect  (or  have  the  potential  to   infect)  humans,  but  a  complex  set  of  organisms  that  display  a  more  stable   association  with  the  human  host.  What  counts  as  a  part  of  the  human  virome   would  then  be  limited  to  viruses  that  have  a  stable  association  with  the  human   body  (be  it  beneficial,  detrimental  or  neutral  with  respect  to  the  survival  of  the   host).     A  third  and  even  more  restricted  understanding  of  the  term  virome   follows  from  a  more  restricted  understanding  of  the  term  microbiome.  A  highly   cited  article  on  the  human  microbiome  project  begins:  “The  human  microbiome   project  (HMP)  reflects  the  fact  that  we  are  supraorganisms  composed  of  human   and  microbial  components”  (Turnbaugh  et  al.,  2007).  Later,  the  same  article   notes:  “The  collective  genomes  of  our  microbial  symbionts  (the  microbiome)   provide  us  with  traits  we  have  not  had  to  evolve  on  our  own”.  This   understanding  of  the  microbiome  appears  specifically  to  exclude  purely  harmful   microbes  –  and  therefore  also  any  harmful  viruses,  if  we  treat  the  virome  as  part   of  the  microbiome.  Crucial  to  this  definition,  then,  is  the  idea  of  ‘good’  or  ‘bad’   microbes.     15   These  examples,  in  summary,  show  different  ways  of  understanding  the   term  ‘human  virome’,  the  first  depending  on  a  distinction  between  (common  and   uncommon)  infections,  the  second  on  the  notion  of  a  ‘stable’  association  between   virus  and  host  and  the  third  depending  on  a  notion  of  the  ‘goodness’  of  a  virus.   Very  probably  there  are  others.   All  of  these  distinctions  have  their  own  problems,  but  the  most  intriguing   is  the  idea  of  a  ‘good’  virus.  As  we  have  discussed  in  Sections  3  and  4,  we  now   know  of  many  cases  where  the  presence  of  the  virus  has  beneficial  effects  on  the   host.  But  does  this  mean  that  there  is  a  set  of  viruses  that  are  good  for  us  and   another  set  that  harm  us?  And  is  this  a  dichotomy  with  which  we  can  define   sharply  the  term  ‘virome’?     In  line  with  what  we  know  about  bacterial  symbiosis,  recent  insights  into   the  roles  played  by  viruses  suggest  that  such  a  classification  is  not  possible.   Whether  a  virus  is  good  or  bad  for  us,  just  as  for  a  bacterium,  is  not  a   characteristic  that  it  carries  like  an  essence.  Whether  a  bacterium  or  a  virus  is   ‘good’  or  ‘bad’  is  always  a  question  of  what  it  does,  and  not  merely  a  question  of   intrinsic  properties.  Some  more  general  philosophical  reflections  on  the   ontological  status  of  viruses  will  help  to  strengthen  this  point.     6.  What  are  viruses?   Viruses  are  extremely  diverse  in  their  structure  and  complexity.  Generally  they   are  fairly  specific  in  their  choice  of  hosts,  and  it  appears  that  just  about  every   kind  of  cell  can  be  host  to  some  kind  of  virus.  Some  viruses  contain  genomes  that   may  code  for  as  few  as  two  proteins  (see,  for  instance,  Nogawa  et  al.,  1996).  At   the  other  extreme  are  the  recently  identified  giant  DNA  viruses  such  as  the     16   mimivirus,  with  genomes  larger  than  those  of  many  bacteria3.  When  we  think  of   the  viruses  that  occupy  our  environment,  as  countable  things  in  the  ocean  or  in   our  bodies,  we  are  probably  thinking  of  virions,  the  stable  state  in  which  viruses   exist  when  they  are  not  actively  engaging  with  cells.  A  virion  contains  a  central   core  of  nucleic  acid,  RNA  or  DNA,  single-­‐  or  double-­‐stranded,  its  genetic  material.   This  is  surrounded  by  a  protein  coat,  or  capsid,  which  both  serves  to  protect  the   viral  genes,  and  has  surface  features  that  enable  it  to  bond  to  appropriate  host   cells.  Some  viruses  also  contain  the  capsid  within  a  phospholipid  bilayer   membrane  that  is  typically  captured  from  the  host  cell  in  the  process  of  exiting   through  the  host  cell  membrane.  Others  contain  a  lipid  layer  directly   surrounding  the  central,  genome-­‐containing,  area.  The  stability  and  inactivity  of   the  virion  lends  intuitive  support  to  the  common  claim  that  viruses  are  not  living   things.       But  the  virion  state  is  neither  the  only  state  in  which  viruses  exist,  nor  is  it   strictly  speaking  inactive  (see  Claverie  &  Abergel;  and  Forterre,  this  issue).  The   virion  can  attach  itself  to  a  suitable  host  cell  where  it  initiates  a  series  of  events   that  typically  lead  to  the  destruction,  or  lysing,  of  the  cell  and  the  production  of   hundreds  or  thousands  of  new  virions.  The  details  of  these  changes  are  again   diverse.  In  some  cases  the  virion  enters  the  cell  complete  with  its  protein  coat,  in   other  cases  this  is  abandoned  and  only  the  genetic  material  enters  the  cell.   Various  molecular  events  may  follow,  but  in  most  cases  the  transcription  and   translation  machinery  of  the  cell  is  eventually  hijacked  for  the  copying  of  the   viral  nucleic  acid  and  the  production  of  viral  proteins.  Some  of  these  proteins                                                                                                                   3  Amazingly,  it  appears  that  there  are  even  viruses  that  attack  these  giant  viruses   (see  LaScola  et  al.,  2008).  The  authors  of  this  report  suggest  that  these   ‘virophages’  might  be  seen  as  vehicles  for  moving  DNA  between  giant  viruses.     17   may  be  involved  in  the  process  of  redirecting  the  activities  of  the  cell’s  genome.   The  virion  in  some  cases  will  self-­‐assemble  automatically,  in  others  it  will   require  the  assistance  of  special  proteins.  In  the  case  of  the  very  large  viruses   such  as  mimivirus,  much  of  the  translation  machinery  is  actually  encoded  by  the   virus.     The  one  thing  that  is  constant  throughout  these  changes  is  the  viral   genome.  One  natural  thought  is  therefore  that  the  virus  should  simply  be   identified  with  its  genetic  material.  Unlike  more  complex  living  systems  there  is  a   stage  in  its  life  cycle  within  the  cell  in  which  it  is,  apparently,  reduced  to  nothing   but  its  genes.  The  genome  might  be  seen  as  the  viral  essence,  with  various   attachments,  notably  the  capsid,  as  contingent  additions  at  various  stages  in  the   life  cycle.  This,  however,  is  not  a  tenable  position  for  several  reasons.     The  first  problem  concerns  the  existence  of  many  viruses  in  a  so-­‐called   latent  state  within  the  host  cell.  In  some  cases,  for  example  of  herpesviruses,  the   latent  state  consists  of  ‘episomes’,  the  viral  DNA  taking  a  circular  form  that  is   then  tethered  to  the  host  chromosome.  In  this  state  the  viral  DNA  must  elude  the   repair  mechanisms  of  the  host  genome,  and  can  undergo  mitosis  alongside  the   host  DNA.  To  this  end  it  undergoes  chromatinization  similar  to  that  of  the  host   chromosome,  and  is  subject  to  epigenetic  modification  (see,  e.g.,  Lieberman,   2008).  Under  these  circumstances,  it  is  questionable  whether  the  viral  genome  is   properly  to  be  seen  as  a  distinct  entity,  rather  than  an  addition,  if  unwelcome,  to   the  host  genome.       This  question  is  more  pressing  when  one  considers  retroviruses,  for   instance  human  immunodeficiency  virus  (HIV).  HIV,  in  the  virion  state,  is  a   single-­‐stranded  RNA  virus.  When  it  invades  a  cell  it  is  reverse  transcribed  into     18   double-­‐stranded  DNA,  and  inserts  itself  into  the  host  genome.  First,  though  the   inserted  DNA  is  isomorphic  to  the  HIV  RNA,  it  is,  nonetheless,  a  different   chemical.  Second,  the  question  whether  inserted  DNA  is  a  distinct  entity  rather   than  an  addition  to  the  host  genome  is  very  hard  to  answer.  If  one  is  tempted  to   insist  that  it  is  a  distinct  entity,  one  might  reflect  that  large  parts  of  eukaryote   genomes  originated  from  viral  insertions  or  intra-­‐genomic  replications  of  viral   genomes,  as  discussed  above  for  the  case  of  ERVs.  For  substantial  periods  after   their  initial  entrapment,  ERVs  may  continue  to  express  themselves  as  viruses,   producing  virions  that  can  infect  other  organisms.  In  time  they  typically  lose  this   ability  and  may  either  become  genomic  junk  or,  frequently,  may  be  adapted  to   serve  vital  functions  in  the  host  organism.  A  number  of  ERVs  are  expressed  in  the   mammalian  placenta,  for  example,  and  appear  to  serve  important  functions,   possibly  exploiting  viral  techniques  for  modifying  host  reactions  to  alien   genomic  elements  (Haig,  2012).  At  what  point  does  an  ERV  cease  to  be  a  virus,   and  become  part  of  the  genome  into  which  it  has  been  inserted?  The  lack  of  any   clear  answer  to  this  question  suggests  that  the  question  is  in  some  way  poorly   posed.       7.  Going  Viral   The  problem  underlying  the  question  just  raised  is  similar  to  the  one  we   encountered  when  trying  to  define  the  virome:  drawing  clear  boundaries  is  not   possible  as  long  as  we  presume  that  a  virus  is  some  sort  of  stable  thing  with   given  properties.  In  the  same  way  that  we  cannot  use  that  picture  to  come  to  an   understanding  of  the  ‘goodness’  or  ‘badness’  of  a  virus,  we  cannot  use  it  to  decide     19   whether  ERV  DNA  is  a  virus  sitting  inside  a  genome  or  not.  Or  so  we  want  to   claim  here.   The  problem  is  that  identifying  the  virus  with  anything  less  than  a  cycle  is   bound  to  lead  to  failure.  A  single  thing,  for  instance  an  episome  or  a  virion,  is   always  less  than  the  actual  virus.  It  is  only  part  of  what  makes  the  virus.  What   matters  is  not  the  DNA  molecule  itself  but  what  it  does  (or  can  do)  in  a  particular   context:  qualifying  as  viral  is  a  matter  of  doing  characteristic  things,  such  as   invading  cells  and  replicating,  rather  than  having  a  particular  intrinsic  property.   This  is  just  as  well,  since  there  are  no  properties  that  a  virus  has  at  every  stage  of   its  life  cycle.     And  this  is  where  the  question  at  the  end  of  the  last  but  one  paragraph   misfires:  an  ERV  is  a  virus  just  as  long  as  it  maintains  the  capacity  to  contribute   to  a  viral  process.  Whether  it  lives  in  a  host  genome  is  immaterial.  In  the  case  of   viral  latency  as  an  episome,  the  episome  should  count  as  viral  because  occupying   that  state,  and  expressing  the  genes  that  it  expresses  to  maintain  itself  during   latency,  are  things  that  some  viruses  do  to  maintain  their  life  cycles.     We  suggest  that  rather  than  trying  to  provide  a  set  of  characteristics  that   qualify  something  as  a  virus,  we  should  concentrate  on  the  activities  that   constitute  the  viral  life  cycle.  In  other  words,  we  should  see  viruses  as  processes   rather  than  things.  Rather  than  worry  whether  an  episome,  or  a  piece  of  naked   DNA  replicating  in  its  host  cell  is  a  virus,  we  should  simply  recognise  that  it  is   part  of  the  viral  process  and  it  is  because  of  this  belonging  to  a  specific  process   that  it  qualifies  as  ‘viral’  DNA.     The  desirability  of  the  process  perspective  is  highlighted  by  the  difficulty   in  counting  viruses.  Here  we  don’t  refer  to  practical  difficulties  (though  these  are     20   considerable!)  but  conceptual  difficulties.  There  is  no  problem  (again   conceptually)  in  counting  virions.  A  single  virion  enters  a  cell  and  at  some  later,   perhaps  much  later,  time,  some  number  of  virions  emerge  from  the  remains  of   the  lysed  cell.  How  many  viruses  were  present  in  the  intervening  time?  Suppose   the  virus  exists  as  multiple  episomes  and  passes  through  several  events  of   mitosis.  Many  of  the  resultant  cells  die  before  the  next  stage  of  viral  replication  is   activated.  All  the  cells  are  infected,  but  are  some  more  infected  than  others,  with   more  viruses,  if  they  have  a  greater  number  of  episomes?  Is  the  death  of  one  of   these  cells  also  the  death  of  a  specific  number  of  viruses?  It  is  not  that  one  could   not  choose  to  attach  answers  to  these  questions  but  it  is  not  clear  why  one  would   want  to.  We  are  interested  in  the  unfolding  of  a  process  rather  than  in  tracing  the   careers  of  discrete  entities,  or  things.     One  final  point  will  conclude  the  present  brief  discussion  on  the  nature  of   viruses.  The  boundaries  between  viruses  and  related  entities  are  not  easy  to   define.  One  very  important  class  of  related  entity  is  the  plasmids,  which  we  have   already  discussed  in  more  detail  in  Section  4.   Plasmids  are  generally  considered  to  be  differentiated  from  viruses  by   their  lack  of  a  capsid;  they  consist  of  naked  DNA.  But  as  we  have  noted,  viruses   do  not  have  capsids  at  all  stages  in  their  life  cycles.  In  fact,  like  viruses,  some   plasmids  can  attach  to  a  host  eukaryotic  genome  in  the  form  of  an  episome.  In   this  state,  indeed,  they  are  not  strictly  naked,  as  they  must  undergo  the   chromatinization  necessary  for  replication  in  mitosis  in  the  host  genome.  So   what  is  the  difference  between  a  viral  and  a  plasmid  episome?  The  answer,   obviously,  is  that  they  are  parts  of  processes  that  differ  at  other  stages  of  their     21   life  cycles.  So  again,  the  only  way  of  understanding  the  generally  intended  limits   of  the  concept  of  a  virus  is  to  recognise  that  a  virus  is  a  process.       8.  Are  viruses  alive?   It  remains  hotly  contested  whether  viruses  should  be  counted  among  the  living   (see  Claverie  &  Abergel;  Forterre;  Koonin;  van  Regenmortel;  Kostyrka,  this  issue).   It  is  not  entirely  obvious  why  this  is  a  question  that  anyone  should  care  deeply   about.  It  sounds  rather  like  either  a  matter  of  fairly  arbitrary  definition,  or  a  relic   of  an  earlier  time  when  being  alive  was  associated  with  some  fundamental   ontological  difference  from  merely  physical  matter.  The  question,  however,  may   help  to  bring  the  different  strands  of  our  discussion  together  and  allow  us  to   further  clarify  the  concept  of  living  systems  as  being  inherently  processual.   Elsewhere  (Dupré  &  O’Malley,  2009)  one  of  us  has  argued  that  standard   reasons  for  denying  that  viruses  are  alive  are  misguided.  In  brief,  the  problem  is   that  most  of  the  criteria  involved  would  exclude  from  the  category  of  the  living   much  that  we  almost  universally  agree  should  be  included.  One  central  example   is  the  criterion  of  autonomy.  Without  going  into  great  detail  about  what  exactly   this  criterion  involves,  the  fact  that  viruses  require  essential  resources  from  the   host  cell  for  their  reproduction  is  often  taken  to  exclude  them  from  the  category   of  the  living.  But  given  that  most  or  all  eukaryotic  organisms,  including  ourselves,   depend  on  a  multitude  of  symbiotic  organisms  for  our  survival  and,  a  fortiori,   reproduction,  it  would  appear  that  on  this  criterion  we  are  not  alive  either— presumably  a  reductio  ad  absurdum  of  the  criterion  in  question.   The  main  thesis  of  that  earlier  paper  is  that  living  systems  are  constituted   by  complex  interactions  between  lineage-­‐forming  elements  of  many  different     22   kinds.  Sometimes,  perhaps  almost  always,  these  elements  include  viruses.  In  that   paper  we  suggested  that  viruses  were  living  when  active  in  these  larger  systems,   but  not  when  in  the  dormant  virion  state.  Van  Regenmortel  (this  issue)  argues   that  a  virus  active  in  a  larger  living  system  is  no  more  a  living  entity  than  any   other  part  of  the  system,  an  organelle  or  an  organ.  It  is  not  clear  whether  a  great   deal  hangs  on  which  side  of  this  debate  one  prefers,  given  that  it  is  doubtful  that   there  is  any  sharp  living/non-­‐living  distinction.  However,  there  is  an  obvious   difference  between  a  virus  and  an  organelle:  a  virus  is  part  of  a  process  distinct   from  the  organism  in  which  it  is  at  a  particular  time  active.  It  has  a  history  that   intersects  that  system  for  a  particular  period  of  time,  and  may  continue  in   another  system.     There  are  two  issues  that  need  to  be  revisited  in  this  discussion  of  the   status  of  viruses,  namely  a  focus  on  collaboration  and  an  emphasis  on  process,   the  two  key  notions  that  we  mentioned  at  the  beginning  of  this  paper.   Collaboration  is  a  prominent  feature  of  any  biological  system,  even  if  no  viruses   are  involved.  Symbiotic  bacteria  are  a  key  example  of  agents  that  work  together   with  other  organisms  to  form  holobionts.  As  a  consequence,  the  actual  agent  (for   instance  the  cow  that  is  said  to  digest  cellulose)  has  now  to  be  understood  as   something  more  complex,  as  it  is  the  larger  symbiotic  system  that  should  be  said   to  do  the  digesting.   We  encounter  the  same  issue  in  the  virus  examples  discussed  above.  We   have  seen  that  in  many  cases  an  organism  cannot  do  what  it  does  without   viruses  being  actively  engaged  inside  it.  And  the  virus  in  turn  cannot  maintain  its   own  cycle  without  the  processes  going  on  inside  the  cell.  But  the  virus  examples   seem  to  take  us  further  than  the  examples  of  symbiotic  bacteria.  If  we  simply     23   look  at  the  interplay  between  bacteria  and  multicellular  organisms  there  might   still  be  a  temptation  to  look  at  what  is  going  on  as  a  collaboration  between  two   stable  things,  each  with  its  own  properties.  Like  a  TV  set  and  a  DVD  player  that   work  together  to  achieve  a  particular  task,  the  bacteria  and  the  host  cell  come   together  as  two  pre-­‐existing  machines  that  are  somehow  compatible  and  work   together.     But  the  example  of  viruses  helps  to  reinforce  the  more  general  point  that   such  a  simple  picture  of  collaboration  has  to  be  expanded.  The  collaboration  we   see  between  virus  and  host  is  not  a  simple  interaction  between  two  stable  things,   but  has  to  be  understood  as  a  collaborative  interaction  between  processes.  As  we   noted  above,  the  virus  itself  can  only  be  understood  if  it  is  depicted  as  a  cycle,   and  the  horizontal  and  lateral  flow  of  genetic  material  has  shown  that  the  host   itself  is  not  a  ‘pure’  and/or  stable  entity  in  itself.  Both  ‘things’  that  come  together   in  the  collaboration  are  intrinsically  fluid  and  temporarily  stabilised  entities   (processes).  And  the  activity  we  are  interested  in  (be  it  digestion  or  the   maintenance  of  a  reproductive  cycle)  turns  out  to  be  a  product  of  a  coming   together  of  different  processes.   In  all  of  the  above  the  coming  together  of  entities  (processes)  has  played  a   key  role.  But  what  comes  together  may  also  separate  again.  We  have  pointed  out,   for  instance,  that  a  key  difference  between  a  virus  and  an  organelle  is  that  the   former  can  re-­‐emerge  from  the  cell  and  continue  its  own  life  cycle.  If  we  adopt   the  perspective  of  a  process  ontology  we  can  make  sense  of  the  constant  merging   and  separating,  because  processes  can  certainly  merge  into  a  single  process  and   may  even  maintain  their  identity  sufficiently  to  part  company  again.  A  slightly   fanciful  example  might  help  to  illustrate  this  point.       24   Imagine  that  two  battles,  both  parts  of  the  same  war  and  hence  involving   the  same  armies,  intersect.  In  each  battle  the  Oceanians  are  driving  the   Eastasians  backwards,  but  in  different  directions.  In  each  case  the  retreating   army  is  following  orders  to  retreat  to  a  particular  rallying  point,  and  their   pursuers  are  under  orders  to  follow  them  in  the  direction  they  are  retreating.   The  battles  cross,  and  continue  in  their  different  directions.  In  the  areas  in  which   they  cross,  however,  the  combatants  are  engaged  in  both  battles.  When  they   encounter  an  enemy  they  will  not,  we  suppose,  enquire  as  to  which  battle  that   soldier  is  fighting  before  deciding  whether  to  engage;  they  are  all  equally   enemies,  to  be  killed  if  possible.  If  soldiers  accidentally  wander  into  a  battle  in   which  their  army  is  engaged,  they  are  liable  to  become  part  of  the  battle.  In  the   case  imagined  this  is  just  what  has  happened  to  all  the  combatants  in  both   battles  who  find  themselves  in  the  zone  of  intersection.   The  point  of  the  example  is  that  it  is  a  teleological  element,  in  this  case  the   intentions  or  orders  of  the  combatants,  that  enables  the  processes  to  re-­‐emerge   from  their  intersections.  By  contrast,  it  is  hard  to  imagine  two  streams  of  water   merging,  flowing  together,  and  then  separating  again,  to  reconstitute  the  very   same  processes  that  were  the  original  merging  flows.  After  the  convergence   there  is  nothing  that  identifies  a  particular  molecule  with  the  flow  from  which  it   originated.  In  the  case  of  the  soldiers  their  intentions  achieve  this  continuity   through  the  period  of  merger.     This,  we  suggest,  is  the  right  way  to  think  of  the  intersection  of  a  virus   with  its  host  organism,  more  specifically  in  the  case  where  the  host  survives  its   interaction  with  the  virus.  We  have  two  processes  which,  for  a  time,  merge  into  a   coherent  process  and  which  may,  thereafter  continue  as  separate  processes  (or,     25   often,  as  one  host  process  and  many  viral  processes).  A  particular  entity,  say  a   viral  episome,  may  be  part  of  more  than  one  process,  for  example  mitosis  as  part   of  the  lineage  of  cells  in  which  it  is  resident,  and  viral  reproduction,  the  process   which  will  separate  off  again  at  some  later  point.     But  viruses  surely  don’t  have  intentions  in  the  same  sense  as  conscious   beings  have  them,  so  how  can  we  claim  that  the  battle  example  is  a  good  guide   for  our  thinking  about  viruses?  It  is  important  to  note  that  in  the  battle  example   the  intention  does  not  merely  describe  an  end  state  at  which  the  soldier   eventually  arrives  somehow  or  other.  The  soldiers  must  keep  their  goals  in  mind   and  allow  these  to  guide  their  actions;  if  the  soldiers  forget  their  orders  then  the   process  will  quickly  break  down.     This  is  important  because  it  brings  the  idea  of  a  process  extended  through   time  into  the  picture.  We  should  not  just  think  of  the  final  goal,  but  also  of  the   beginning  of  the  process  and  the  activities  that  tie  together  its  various  stages;   where  one  is  coming  from  matters  as  much  as  where  one  is  going.  The  same  view,   we  claim,  applies  to  the  virus  example.  A  virus  does  not,  of  course,  have  a   memory  any  more  than  it  has  intentions.  But  there  nevertheless  is  a  certain   coherence  and  interconnectedness  between  the  different  activities  that  we  label   ‘viral’  and  which  we  take  to  constitute  a  viral  life  cycle.  The  genes  an  episome   expresses,  for  instance,  are  expressed  to  maintain  the  viral  life  cycle.  The   function  of  the  episome  is  therefore  not  just  defined  relative  to  the  actual  system   in  which  it  finds  itself  (a  mitotic  cell  for  example),  but  it  is  related  to  the  future   and  the  past  of  the  virus  (thought  of  as  an  active  cycle).  We  cannot  understand   what  makes  the  episome  a  viral  DNA  if  we  don’t  take  into  account  the  whole     26   cycle  or  process  that  is  the  virus.  In  contrast,  a  mere  DNA  plasmid  is  not  working   as  part  of  this  viral  cycle  and  this  is  where  it  clearly  differs  from  a  viral  episome.   This  nicely  resonates  with  how  Nicholas  Rescher  describes  processes   (Rescher,  1996).  Rescher  introduces  a  view  of  processes  as  being  defined  by  a   functional  unity;  there  is  a  ‘programmatic  structure’  that  characterises  and   unifies  a  process.  Interconnected  activities  that  form  a  functional  unity  are  key  to   his  understanding  of  processes:  “A  process  is  made  into  the  item  it  is  not  through   its  continuing  (“essential”)  properties,  as  with  a  classically  conceived  substance,   but  by  its  history,  by  the  temporal  structure  of  its  descriptive  unfolding  across   time.  The  identity  of  a  process  is  constituted  through  a  sequential  pattern  of   action  […]”  (Rescher,  1996,  p.41).4     We  see  such  a  functional  unity  also  in  the  case  of  viruses  but  not  in  the   case  of  the  two  water  streams  that  combine  and  separate  again.  In  the  case  of  the   two  intersecting  streams  of  water  we  do  of  course  have  some  form  of  structure,   the  physical  boundaries  that  help  us  identify  the  two  streams.  But  nothing  in  this   structure  makes  the  streams  reform  after  the  intersection.  The  viral  life  cycle  has   a  stronger  unity  than  this,  as  it  manages  to  intersect  with  other  processes  and  re-­‐ emerge  again.   What  gives  the  process  ‘its’  structure  is  a  difficult  question.  What  is  clear   in  the  case  of  the  virus-­‐as-­‐process  is  that  it  cannot  simply  be  the  DNA  (or  any  of   the  other  entities  that  are  part  of  the  viral  cycle),  because  we  have  seen  that  the   entity  captured  by  the  idea  ‘virus’  is  much  larger  than  simply  its  genome  or  other                                                                                                                   4  Another  approach  to  this  problem  is  through  Reichenbach’s  concept  of   genidentity,  as  discussed  in  (Guay  &  Pradeu,  2016).     27   specific  stages;  it  is  always  a  whole  cycle,  i.e.  a  running  process  or  overarching   activity,  that  defines  the  virus.   This  process-­‐centred  perspective  provides  a  very  different  understanding   of  activity  and  function  in  biological  systems  from  merely  the  interaction  of   discrete  individual  things,  a  difference  that  is  ultimately  evolutionary.  The   episome  is  part  of  a  process  that  has  a  pattern  (the  cycle)  and  it  is  this  cycle  that   has  evolved  its  distinctive  trajectory.  The  interconnectedness  between  different   parts  of  such  processes  has  developed  over  evolutionary  time  and  is  what  unites   these  stages  into  the  larger  whole  that  we  refer  to  as  the  virus.  It  is  in  this   coherence  that  we  find  the  analogy  to  the  intention  and  memory  that  determine   the  soldier’s  continued  participation  in  a  particular  battle.  At  the  same  time,  the   persistence  of  this  process  has  come  to  depend  on  multiple  interactions  with   further  living  processes  that  have  each  their  own  cyclical  structure  and  mode  of   persistence.  Crucially,  the  virus  may  both  have  its  independent  cyclical  structure   and  be  a  necessary  part  of  a  larger  system  with  which  it  interacts.  Hence  there   need  be  no  unique  division  of  biological  activity  into  distinct,  non-­‐overlapping   processes.   The  notion  of  the  independent  active  cycle  that  has  evolved  over  time  also   brings  us  back  to  the  living  vs.  not-­‐living  question.  We  do  not  suggest  that  an   entity  is  living  simply  by  virtue  of  being  part  of  a  living  system.  What  more  is   required?  The  key  is,  again,  to  recognise  the  essentially  processual  and   interconnected  character  of  life.  We  should  not  try  to  decide  whether  an  entity,  a   traditional  substance,  has  a  set  of  properties  that  qualify  it  as  living,  but  rather  to   identify  activities  that  sustain  both  conceptually  and  actually  separable   processes  in  the  characteristic  and  extraordinarily  efficacious  ways     28   characteristic  of  life  as  we  know  it.  If  we  simply  consider  viruses  as  particles  then   there  is  no  doubt  that  they  are  seriously  deficient  in  the  qualities  needed  to   sustain  living  processes.  On  the  other  hand,  as  we  have  seen  above,  viruses  do   have  the  capacity  to  contribute  powerfully  to  such  sustenance.  The  key  to   reconciling  these  observations  is  once  again  to  adopt  the  processual  perspective.   It  is  the  viral  process  (and  not  simply  a  virus  particle)  that  can  take  on  a   functional,  or  even  essential,  role  for  the  processes  it  intersects  with  (for   instance  by  regulating  bacterial  cell  numbers  (Section  3)).5  And  as  some  of  these   latter  processes  are  unquestionably  living  processes,  for  example  ourselves,  it   would  seem  odd  to  deny  the  status  of  living  to  the  viral  processes  that  interact   with  them  in  this  life-­‐sustaining  way.  Virions,  then,  are  not  living  things  certainly,   but  they  are  stages  of  living  processes.       9.  Conclusion   Viruses  raise  questions  on  many  different  levels.  There  is,  for  instance,  the   question  of  whether  they  are  good  or  bad  for  us,  as  figured  in  the  discussion  of   the  term  ‘virome’.  There  is  also  the  question  of  whether  viruses  belong  to  the   realm  of  living  things,  with  which  we  ended  this  paper.  And  there  is  of  course  the   issue  of  what  a  virus  actually  is.     For  all  of  these  issues  a  substance  ontology  has  little  to  offer.  If  we  take   things  and  their  fixed  set  of  properties  to  be  fundamental  we  find  it  hard  to  come   to  any  answers  to  these  various  problems  of  classification.  One  reason  for  this  is                                                                                                                   5  We  think  in  general  that  the  grounding  of  capacities  in  relational  and  dynamic   aspects  of  entities  or  systems  is  an  important  feature  that  a  processual   perspective  helps  to  emphasise  and  articulate  (for  further  discussion  see  also   (Guttinger,  forthcoming)).     29   that  the  substance  standpoint  assumes  an  essentialism  and/or  individualism,   neither  of  which  squares  well  with  the  interconnected  picture  of  the  biological   world  that  the  natural  sciences  are  painting  for  us.   Switching  to  a  focus  on  processes  allowed  us  to  make  sense  of  distinctions   that  are  often  used  in  the  natural  sciences,  for  instance  the  distinctions  between   a  plasmid  and  a  viral  episome  or  between  a  good  and  a  bad  virus.  What  matters   in  both  these  cases  is  that  the  ‘thing’  we  are  talking  about  is  a  process  that  can   often  mix  with  and  become  part  of  other  processes  and  hence  contribute  to  a   range  of  outcomes  at  the  same  time.  What  keeps  the  viral  process  separate  from   the  other  processes  with  which  it  is  intertwined  is  the  interconnectedness  that   its  different  sub-­‐processes  display  and  to  which  we  have  referred  above  using   Rescher’s  term  of  ‘functional  unity’.   Clearly  more  work  is  needed  to  further  develop  this  idea  of  a  functional   unity.  How  is  it  brought  about  and  how  is  it  maintained?  On  what  basis  do  we   attribute  it  to  some  part  of  the  biological  flux,  and  are  we  even  sure  what  it  is   that  we  are  attributing?     But  even  though  there  still  are  many  open  questions  related  to  this  unity,   there  are  also  some  preliminary  insights  we  can  take  away  from  the  above   discussion  of  viruses.  One  insight  is  related  to  the  question  of  what  provides  the   unity  we  observe  in  the  case  of  viruses  but  not  the  intersecting  streams  of  water.   If  we  could  find  an  essential  property  that  defined  the  virus,  the  problem  would   be  easily  solved.  But  the  only  candidate  for  such  a  property  is  the  genome  and  as   we  have  argued,  the  genome  cannot  serve  this  purpose.  As  we  have  seen,  the   viral  life-­‐cycle  should  not  be  understood  as  some  sort  of  material  thing  or  as  a   mere  succession  of  different  states  of  one  material  thing  that  gives  the  cycle  its     30   unity.  Rather,  the  process  of  the  cycle  as  a  whole  is  the  virus.  And  the  viral  life-­‐ cycle  is  one  of  the  many  processes  that  may  come  together  to  form  yet  another   stable  pattern  to  which  we  usually  refer  as  ‘organism’.       Acknowledgments:   We  thank  Thomas  Pradeu  and  Ann-­‐Sophie  Barwich  for  very  helpful  comments   and  suggestions  on  an  earlier  draft.    We  are  also  grateful  to  the  members  of  the   ProBio  discussion  group  at  the  Egenis  Centre  for  insightful  discussion.  The   research  leading  to  this  paper  has  received  funding  from  the  European  Research   Council  under  the  European  Union’s  Seventh  Framework  Programme   (FP7/2007-­‐2013)  /  ERC  grant  agreement  n°  324186.       31   Bibliography   Bapteste,  E.,  O’Malley,  M.  A.,  Beiko,  R.  G.,  Ereshefsky,  M.,  Gogarten,  J.  P.,  Franklin-­‐ Hall,  L.,  Lapointe,  F.  J.,  Dupré,  J.,  Dagan,  T.,  Boucher,  Y.,  &  Martin,  W.  (2009).   Prokaryotic  evolution  and  the  tree  of  life  are  two  different  things.  Biology  Direct,   4(34),  1-­‐20.     Barr,  J.  J.,  Auro,  R.,  Furlan,  M.,  Whiteson,  K.  L.,  Erb,  M.  L.,  Pogliano,  J.,  Stotland,  A.,   Wolkowicz,  R.,  Cutting,  A.  S.,  Doran,  K.  S.,  Salamon,  P.,  Youle,  M.,  &  Rohwer,  F.   (2013).  Bacteriophage  adhering  to  mucus  provide  a  non–host-­‐derived  immunity.   Proceedings  of  the  National  Academy  of  Sciences,  110(26),  10771-­‐10776.     Bouchard,  F.,  &  Huneman,  P.  (eds.).  (2013).  From  Groups  to  Individuals:  Evolution   and  Emerging  Individuality.  MIT  Press.     Bravo,  J.  A.,  Julio-­‐Pieper,  M.,  Forsythe,  P.,  Kunze,  W.,  Dinan,  T.  J.,  Bienenstock,  J.,  &   Cryan,  J.  F.  (2012).  Communication  between  gastrointestinal  bacteria  and  the   nervous  system.  Current  Opinion  in  Pharmacology,  12,  667–672.     Brüssow,  H.,  &  Hendrix,  R.  W.  (2002).  Phage  Genomics:  Small  Is  Beautiful.  Cell,   108,  13–16.     Chu,  H.,  &  Mazmanian,  S.  K.  (2013).  Innate  immune  recognition  of  the  microbiota   promotes  host-­‐microbial  symbiosis.  Nature  Immunology,  14(7),  668-­‐75.       32   Delwart,  E.  (2013).  A  Roadmap  to  the  Human  Virome.  PLoS  Pathogens,  9(2),   e1003146.     Dupré,  J.  (2005).  Are  There  Genes?  In  A.  O’Hear  (Ed.),  Philosophy,  Biology  and  Life   (Royal  Institute  of  Philosophy  Supplements)  (pp.  193-­‐210).  Cambridge:  Cambridge   University  Press.     Dupré,  J.  (2012).  Processes  of  Life:  Essays  in  the  Philosophy  of  Biology.  Oxford:   Oxford  University  Press.     Dupré,  J.,  &  O'Malley,  M.  A.  (2009).  Varieties  of  living  things:  life  at  the   intersection  of  lineage  and  metabolism.  Philosophy  &  Theory  in  Biology,  1.     Griffiths,  D.  J.  (2001).  Endogenous  retroviruses  in  the  human  genome  sequence.   Genome  Biology,  2,  1017.     Guay,  A.,  &  Pradeu,  T.  (eds.)  (2016).  Individuals  Across  the  Sciences.  New  York:   Oxford  University  Press.     Guttinger,  S.  (forthcoming).  Towards  a  Process  Ontology  for  Macromolecular   Biology.  In  J.  Dupré  &  D.  Nicholson  (Eds.),  Everything  Flows:  Towards  a  Processual   Philosophy  of  Biology.  Oxford  University  Press.       33   Guttinger,  S.,  Laurell,  E.,  &  Kutay,  U.  (2009).  Orchestrating  nuclear  envelope   disassembly  and  reassembly  during  mitosis.  Nature  Reviews  Molecular  Cell   Biology,  10(3),  178-­‐91.       Haig,  D.  (2012).  Retroviruses  and  the  Placenta.  Current  Biology,  22(15),  R609  -­‐   R613.     Hehemann,  J.  H.,  Correc,  G.,  Barbeyron,  T.,  Helbert,  W.,  Czjzek,  M.,  &  Michel,  G.   (2010).  Transfer  of  carbohydrate-­‐active  enzymes  from  marine  bacteria  to   Japanese  gut  microbiota.  Nature,  464(7290),  908-­‐912.     Henning,  B.  G.  (2013).  Of  Termites  and  Men:  On  the  Ontology  of  Collective   Individuals.  In  B.  G.  Henning,  &  A.  C.  Scarfe  (Eds.),  Beyond  Mechanism:  Putting  Life   Back  Into  Biology  (pp.  233  –  248).  Plymouth:  Lexington  Books.     Lander,  E.  S.,  Linton,  L.  M.,  Birren,  B.  et  al.  (2001).  Initial  sequencing  and  analysis   of  the  human  genome.  Nature,  409(6822),  860–921.     La  Scola,  B.,  Desnues,  C.,  Pagnier,  I.,  Robert,  C.,  Barrassi,  L.,  Fournous,  G.,  Merchat   M.,  Suzan-­‐Monti,  M.,  Forterre,  P.,  Koonin,  E.,  &  Raoult.  D.  (2008).  The  virophage   as  a  unique  parasite  of  the  giant  mimivirus.  Nature,  455(7209),  100-­‐104.     Lecuit,  M.,  &  Eloit,  M.  (2013).  The  human  virome:  new  tools  and  concepts.  Trends   in  Microbiology,  21(10),  510-­‐515.       34   Lieberman,  P.  M.  (2008).  Chromatin  Organization  and  Virus  Gene  Expression.   Journal  of  Cellular  Physiology,  216(2),  295–302.       Méthot,  P.  O.,  &  Alizon,  S.  (2014).  What  is  a  pathogen?  Towards  a  process  view  of   host-­‐parasite  interactions.  Virulence,  5(8),  775-­‐785.     Mindell,  D.  P.  (1992).  Phylogenetic  consequences  of  symbioses:  Eukarya  and   Eubacteria  are  not  monophyletic  taxa.  BioSystems,  27,  53-­‐62.     Minot,  S.,  Sinha,  R.,  Chen,  J.,  Li,  H.,  Keilbaugh,  S.  A.,  Wu,  G.  D.,  Lewis  ,  J.  D.,  &   Bushman,  F.  D.  (2011).  The  human  gut  virome:  Inter-­‐individual  variation  and   dynamic  response  to  diet.  Genome  Research,  21(10),  1616-­‐1625.     Noble,  D.  (2006).  The  Music  of  Life:  Biology  beyond  the  Genome.  Oxford:  Oxford   University  Press.     Nogawa,  M.,  Kageyama,  T.,  Nakatani,  A.,  Taguchi,  G.,  Shimosaka,  M.,  &  Okazaki,  M.   (1996).  Cloning  and  Characterization  of  Mycovirus  Double-­‐stranded  RNA  from   the  Plant  Pathogenic  Fungus,  Fusarium  solani  f.  sp.  Robiniae.  Bioscience,   Biotechnology,  and  Biochemistry,  60(5),  784-­‐788.   Philippot,  L.,  Raaijmakers,  J.  M.,  Lemanceau,  P.,  &  van  der  Putten,  W.  H.  (2013).   Going  back  to  the  roots:  the  microbial  ecology  of  the  rhizosphere.  Nature  Reviews   Microbiology,  11(11),  789-­‐99.   Pradeu,  T.,  &  Carosella,  E.  D.  (2006).  The  Self  Model  and  the  Conception  of     35   Biological  Identity  in  Immunology.  Biology  and  Philosophy,  21(2),  235-­‐252.   Rescher,  N.  (1996).  Process  Metaphysics:  An  Introduction  to  Process  Philosophy.   Albany:  State  University  of  New  York  Press.     Reyes,  A.,  Haynes,  M.,  Hanson,  N.,  Angly,  F.  E.,  Heath,  A.  C.,  Rohwer,  F.,  &  Gordon,  J.   I.  (2010).  Viruses  in  the  faecal  microbiota  of  monozygotic  twins  and  their   mothers.  Nature,  466(7304),  334-­‐338.   Rohwer,  F.,  Seguritan,  V.,  Azam,  F.,  &  Knowlton,  N.  (2002).  Diversity  and   distribution  of  coral-­‐associated  bacteria.  Marine  Ecology  Progress  Series,  243,  1– 10.   Rosenberg,  E.,  Koren,  O.,  Reshef,  L.,  Efrony,  R.,  &  Zilber-­‐Rosenberg,  I.  (2007).  The   role  of  microorganisms  in  coral  health,  disease  and  evolution.  Nature  Reviews   Microbiology,  5,  355-­‐362.   Round,  J.  L.,  &  Mazmanian,  S.  K.  (2009).  The  gut  microbiota  shapes  intestinal   immune  response  during  health  and  disease.  Nature  Reviews  Immunology,  9(5),   313-­‐23.   Sherratt,  D.  J.  (1974).  Bacterial  Plasmids.  Cell,  3,  189-­‐195.     Smillie,  C.,  Garcillán-­‐Barcia,  M.  P.,  Francia,  M.  V.,  Rocha,  E.  P.,  de  la  Cruz,  F.  (2010).   Mobility  of  plasmids.  Microbiol  and  Molecular  Biology  Reviews,  74(3),  434-­‐52.       36   Spasova,  D.  S.,  &  Surh  C.  D.  (2014).  Blowing  on  embers:  commensal  microbiota   and  our  immune  system.  Frontiers  in  Immunology,  5,  318.   Suttle,  C.  A.  (2007).  Marine  viruses  —  major  players  in  the  global  ecosystem.   Nature  Reviews  Microbiology,  5,  801-­‐812.   Turnbaugh,  P.  J.,  Ley,  R.  E.,  Hamady,  M.,  Fraser-­‐Liggett,  C.  M.,  Knight,  R.,  &  Gordon,   J.  I.  (2007).  The  human  microbiome  project.  Nature,  449(7164),  804-­‐810.   Van  Regenmortel,  MHV.  (2015).  The  metaphor  that  viruses  are  living  is  alive  and   well.  This  issue.     Whitman,  W.  B.,  Coleman,  D.  C.,  &  Wiebe,  W.  J.  (1998).  Prokaryotes:  The  unseen   majority.  Proceedings  of  the  National  Academy  of  Sciences,  95(12),  6578-­‐6583.     Wylie,  K.  M.,  Weinstock,  G.  M.,  &  Storch,  G.A.  (2012).  Emerging  View  of  the   Human  Virome.  Translational  research  :  the  journal  of  laboratory  and  clinical   medicine,  160(4),  283-­‐290.     37