Supplemental material to this article can be found at:

http://dmd.aspetjournals.org/content/suppl/2015/06/01/dmd.115.063602.DC1.html
1521-009X/43/8/1236–1245$25.00 http://dx.doi.org/10.1124/dmd.115.063602
DRUG METABOLISM AND DISPOSITION Drug Metab Dispos 43:1236–1245, August 2015
Copyright ª 2015 by The American Society for Pharmacology and Experimental Therapeutics

A Novel Loading Method for Doxycycline Liposomes for Intracellular

Drug Delivery: Characterization of In Vitro and In Vivo Release
Kinetics and Efficacy in a J774A.1 Cell Line Model of Mycobacterium
smegmatis Infection s
Rebekah K. Franklin, Sarah A. Marcus, Adel M. Talaat, Butch K. KuKanich, Ruth Sullivan,
Lisa A. Krugner-Higby, and Timothy D. Heath
Pharmacology, Clinical, Analytical, and Toxicological Services and the Department of Anatomy and Physiology, Kansas State
University, Manhattan, Kansas (B.K.K.); and Departments of Surgical Sciences (L.A.K.-H.), Pathobiological Sciences (R.S.), and
Comparative Biosciences (A.M.T., S.A.M.); School of Veterinary Medicine (R.S.); School of Pharmacy (T.D.H.); and Research
Animal Resources Center, University of Wisconsin, Madison, Wisconsin (R.K.F., L.A.K.-H., R.S.)
Received January 29, 2015; accepted June 1, 2015

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

ABSTRACT
Doxycycline (doxy) is used in treating intracellular and extracellular STD-doxy, and serial blood samples were collected. Pharmacoki-
infections. Liposomal (LE) antibiotics allow low-frequency dosing netics were analyzed using high-performance liquid chromatogra-
and extended efficacy compared with standard (STD) formulations. phy. Liver and injection site skin samples were collected at
We developed a novel sulfuric acid–loading method for doxycycline euthanasia (4 weeks postinjection). Minimal histologic tissue
liposomes (LE-doxy). We hypothesized that a single s.c. injection of reactions occurred after injection of STD (nonliposomal), DPPC,
LE-doxy would be detectable in serum for at least 2 weeks at or sphing-doxy. DPPC-doxy had slightly faster in vitro leakage than
concentrations equal to or better than STD-doxy and would be sphing liposomes, although both were detectable at 264 hours. The
bactericidal in an in vitro Mycobacterium smegmatis infection of mean residence time for DPPC was the highest (111.78 hours),
J774A.1 macrophage cells. Liposomes were encapsulated by followed by sphing (56.00 hours) and STD (6.86 hours). DPPC and
sulfuric acid gradient loading, and release kinetics were performed sphing-doxy were detectable at 0.2 mg/ml in serum at 336 hours
in vitro and in vivo. LE-doxy made using 8.25 mg/ml doxycycline postadministration. LE-doxy was not toxic to J774A.1 cells in vitro
loaded for 24 hours achieved 97.77% capture in 1,2-dipalmitoyl-sn- and produced inhibition of viable Mycobacterium smegmatis at 24
glycero-3-phosphocholine (DPPC) and 43.87% in sphingomyelin and 48 hours. LE-doxy will require further testing in in vivo infection
(sphing). Rats were injected s.c. with 50 mg/kg LE-doxy or 5 mg/kg models.

Introduction and the antibiotic affinity for macrophages (Denis et al., 1990; Henao
Doxycycline (doxy) is a second-generation tetracycline antibiotic et al., 2007). Often, all of these bacterial infections require long-term,
approved for use in the United States since 1967 (Nelson and Levy, daily treatments, which may lead to poor compliance and under-
2011). Doxy is a broad spectrum antibiotic used to treat a variety of dosing with currently available formulations. The development of new
human and animal diseases. Doxy is effective against bacteria, such as delivery vehicles is important to both human and veterinary medicine.
Coxiella burnetii and Borellia spp. (Nelson and Levy, 2011). It can be An ideal drug can be given at infrequent intervals, yet maintain a high
used to treat intracellular infections as diverse as Plasmodium level of efficacy during the treatment period. Both free and liposomal
falciparum, Chlamydia spp., and Mycoplasma spp., for which it is tetracycline and doxy have been explored in in vitro and in vivo
the drug of choice (Rolain et al., 2005; Zeidner et al., 2008; Nelson models for their antichlamydial activities (Sangaré et al., 1998, 1999,
and Levy, 2011). Bacteria-infecting cells, such as macrophages, are 2001a,b; Selliah and Ravaoarinoro, 2004). Extended-release formula-
challenging to treat due to the location of macrophages in the tissue tions of antibiotics have been used in veterinary medicine for many
years. For example, an oxytetracycline formulation, Liquamycin LA-
200 (Zoetis, Inc., Kalamazoo, MI), which can be administered every
48 hours, has been used in cattle (Kumar and Malik, 1998). However,
This work was supported in part by departmental funds from the Research
a particular formulation may not have the same efficacy across
Animal Resources Center, University of Wisconsin, Madison, Wisconsin; the
Companion Animal Fund, School of Veterinary Medicine, University of Wisconsin,
species. Cefovecin, Convenia (Zoetis, Inc.), a cephalosporin formu-
Madison, Wisconsin; a gift from Comfort Care for Animals, LLC; and the National lation that lasts for 2 weeks, has found wide use in companion animals
Institutes of Health [Grant NIH-1R21AI090308]. in the management of chronic dermatologic diseases. However,
dx.doi.org/10.1124/dmd.115.063602. Convenia relies on resorption of the drug in the renal proximal tubule
s This article has supplemental material available at dmd.aspetjournals.org. in dogs and cats for its extended action, as opposed to liposomal

ABBREVIATIONS: AUC, area under the curve; doxy, doxycycline; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; sphing, sphingomyelin;
LE, liposomal; LE-doxy, liposomal encapsulated doxycycline; MRT, mean residence team; SPE, solid phase extraction; sphing, sphingomyelin;
STD, standard.

1236
 Loading Liposome Doxycycline Pharmacokinetics Efficacy 1237
formulations, which would be expected to have an extended release of a concentration of either 33 or 8.25 mg/ml, and 250 ml was added to the lipid
the drug (Stegemann et al., 2006a,b). Nonhuman primates do not swelled in sulfuric acid, followed by 53.6 ml of 5 M NaOH in 1 M citrate
reabsorb this drug in the kidney, and its half-life is not extended in buffer. The addition of NaOH/citrate raised the pH of the extra-liposomal
these species (Raabe et al., 2011). compartment to pH 3.9, creating a pH gradient that drove the loading of the
doxy. The maximum pH was determined experimentally, and pH 3.9 was
In human medicine, tetracyclines are also used to treat skin
selected because higher pH values resulted in precipitation of the doxy. A
disorders that do not necessarily have a bacterial origin (Gu et al., single tube of doxy liposomes was loaded either on a rotating shaker for
2012). There are formulations of extended-release doxycycline hyclate 24 hours at 22C or in a water bath at 55C for 1 hour. The liposomes were
for human use, but they have not been developed or dosed for their diluted in 4 ml of normal saline (Baxter International, Deerfield, IL),
bacteriocidal activities (Sapadin and Fleischmajer, 2006; Gu et al., sedimented at 1000g for 10 minutes in a centrifuge, and resuspended in 2 ml of
2012). There is evidence that doxy has anti-inflammatory properties at normal saline. The loading efficiency was determined by solubilizing the
subantimicrobial doses and may cause inhibition of phospholipase A2, aqueous liposome suspension in chloroform/methanol at a volumetric ratio of
inhibition of the expression of nitric oxide synthetase, accelerated 1:3:1 water/methanol/chloroform.
degradation of nitric oxide synthase, inhibition of pathologic matrix Particle Size. The particle size for the liposomes was determined using
metalloproteinase activity, and free radical scavenging (Sapadin and a Malvern Nano Series Zetasizer (Malvern Instruments, Malvern, UK). An
aliquot of the liposomal preparation was suspended in 0.9% w/v NaCl. The
Fleischmajer, 2006; Federici, 2011; Gu et al., 2012). These activities
samples were read at 25C.
are used to treat multifactorial inflammatory diseases, such as acne and Thin Layer Chromatography. To determine whether the preparation of
rosacea, in human beings (Sapadin and Fleischmajer, 2006; Federici, liposomes by acid loading led to any degradation of the lipid components,
2011; Gu et al., 2012). a sample of liposomes prepared from DPPC and cholesterol was extracted
Liposomes have long been studied as a drug delivery system (Hunt

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

using the method of Bligh and Dyer (1959). Silica gel 60 thin layer
et al., 1979). Variations in liposome composition, formation, and chromatography plates (250 m) (EMD Millipore, Billerica, MA) were first
loading methodology may lead to strikingly different drug concen- activated by heating on a hot plate for 2 minutes. The extracted lipid was
trations and release kinetics for a particular drug (Hunt et al., 1979; spotted on the plate together with standards for cholesterol and DPPC. The
Omri et al., 1995; Omri and Ravaoarinoro, 1996; Sangaré et al., 1998; plate was developed in a mixture of chloroform/methanol/ammonium
Krugner-Higby et al., 2003, 2009; Smith et al., 2003, 2013; Budai hydroxide 65:25:4 v/v/v. The developed plate was dried and stained with
iodine vapor.
et al., 2009). Our laboratory has developed a novel liposome loading
In Vitro Release Kinetics of LE-Doxy Formulations. Cellulose dialysis
methodology (Heath et al., 2014). We hypothesized that formulations tubing (Fisher Scientific, Pittsburgh, PA) was prepared by boiling in three
of liposomal doxy (LE-doxy) prepared using this technology would be changes of EDTA and sodium carbonate. The tubing was rinsed in distilled
detectable in serum after a single s.c. injection for at least 2 weeks at water and stored in 30% ethanol at 5C until use, as previously described
concentrations equal to or better than that of standard (STD) injectable (Krugner-Higby et al., 2003). Sections of tubing were tied at one end and filled
doxy. We hypothesized that LE-doxy would be bacteriocidal and with 100 ml of DPPC/cholesterol LE-doxy preparations made using 33 or
preserve cell viability in an in vitro model of Mycobacterium 8 mg/ml doxy solutions and 900 ml of normal saline. The ends were tied off,
smegmatis infection of murine origin J774A.1 macrophage line cells. and the bag was placed in 9.0 ml of saline in a 50-ml conical centrifuge tube.
Tubes were prepared in duplicate, covered with tin foil, and placed on a rotary
shaker at 22C. At selected time points, an aliquot of buffer was removed,
Materials and Methods placed in a cuvette, and then absorbance at 345 and 480 nm was determined by
Animals. The University of Wisconsin-Madison School of Veterinary a spectrophotometer. The aliquot of the buffer was returned to the centrifuge
Medicine Animal Care and Use Committee approved the animal protocol. tube, and the tubes were covered with foil and returned to the shaker until the
Experiments were performed on 17 adult August Copenhagen Irish (ACI) male next time point.
and female rats. Sentinel rats in the same facility tested free of ecto- and endo- Blood Draws. Rats were placed under isoflurane anesthesia and kept warm
parasites and were serologically negative for titers to the following antigens: on a reheatable gel pad covered with surgical drape. A 3-cc syringe and a 22-
Mycoplasma pulmonis, Sendai, sialodacryoadenovirus, rat corona virus, gauge needle were used to draw 0.5-ml blood samples from the ventral tail
Kilham virus of rats, Toolans H-1 virus, rat parvovirus, and Clostridium artery. Blood was placed into a 1.5-ml microcentrifuge tube containing glass
piliforme. shards to provide a nidus for clot formation and retraction and was centrifuged.
Rats were housed in polypropylene caging and were provided a commercial Serum was separated from formed elements and frozen at 220C until assayed.
rat diet (Teklad, 7002, 6% fat mouse and rat diet; Harlan, Madison, WI) and Blood was drawn prior to injection from each of the 17 rats. Blood samples
water ad libitum. Rats were weighed at the beginning of the experiment. All were drawn at 1 hour and 4, 8, and 24 hours in the rats administered STD-doxy.
blood collection and injection of drugs were performed under isoflurane anesthesia Blood samples were drawn from the rats administered LE-doxy in DPPC or
administered via facemask to effect. Four weeks after drug injection, rats were sphing at 1 hour and 4, 24, 48, 72, 96, 168 (7 days), and 336 (14 day) hours
deeply anesthetized with isoflurane and euthanized with an intracardiac injection of after injection.
commercially available euthanasia solution (Fatal Plus; Vortech Pharmaceuticals, Injections. STD-doxy was dosed s.c. at 5 mg/kg, and LE-doxy (DPPC or
Dearborn, MI). Samples of liver and the injected rear legs were collected. Liver sphing) was dosed at 50 mg/kg and given s.c. using a 22-gauge needle. The rat
samples were frozen at 220C until evaluated for doxy concentration, and rear was placed under isoflurane anesthesia, and the upper thigh and hip area was
legs were marked with an identification tag and placed in 10% neutral buffered clipped and swabbed with alcohol. Rats were allowed to recover from
formalin until histopathology was performed. anesthesia and were monitored until awake. The rats were observed for any
Liposome Encapsulation by Sulfuric Acid Gradient Loading. A mixture signs of discomfort in the injected site (abnormal gait, shaking, or other
of 80 mmol 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) or 80 mmol abnormal movements).
sphingomyelin (sphing) and 40 mmol cholesterol (Avanti Polar Lipids, Assay for In Vivo Pharmacokinetics of LE-Doxy. Serum samples were
Alabaster, AL) was dried by removing chloroform, dissolved in 1 ml of sterile analyzed for doxy using high-pressure liquid chromatography (Shimadzu
tert-butanol (Sigma-Aldrich, St. Louis, MO) by heating to 55C in a water bath, Prominence; Shimadzu Scientific Instruments, Columbia, MD) with triple
frozen in a dry ice–isopropanol mixture, and lyophilized for 24 hours. The quadruple mass spectrometry (API 2000; Applied Biosystems, Foster City,
resulting microporous lipid masses were stored at 220C until use. One CA). A protein precipitation method was used for sample processing, in which
milliliter of 3.0 M sulfuric acid (AMEND Drug and Chemical Company, a 0.1-ml sample or standard was added to a 0.4-ml internal standard solution
Irvington, NJ), was added to the lipid mass and incubated for 30 minutes at 50C. (oxytetracycline 0.5 mg/ml in methanol with 0.1% formic acid). The processed
A 250-ml aliquot was dispensed into each of the four separate tubes. samples were vortexed for 5 seconds and then centrifuged for 5 minutes at
Doxycycline hyclate (Sigma-Aldrich) was dissolved in sterile water at 15,000g. The supernatant was transferred to an injection vial. The mobile phase
1238 Franklin et al.

consisted of 0.1% formic acid in deionized water (A) and acetonitrile (B). The heavy that individual colonies could not be identified. The microplate Alamar
mobile phase gradient started at 100% A from 0 to 0.5 minutes, with a linear blue assay was used to determine the efficacy of STD and DPPC LE-doxy in
gradient to 40% A at 3.5 minutes, followed by a linear gradient to 100% A at a cell-free system as a comparison (Franzblau et al., 1998).
4.5 minutes, with a total run time of 6.5 minutes. Separation was achieved at Statistical Analysis. The Mann-Whitney test was used to compare the
40C with a 50  2.1 mm 5 mM column (Allure PFP Propyl; Restek serum and liver concentrations of doxy and to compare cell counts and numbers
Corporation, Bellefonte, PA). The qualifying and quantifying ions for doxy of colony-forming units between and within groups. Fisher’s exact test was
were mass to charge (m/z) 445 and 428, respectively. The qualifying and used to compare the probability of histologic lesions in skin samples obtained
quantifying ions (m/z) for oxytetracycline were 461 and 426, respectively. from the rats at necropsy.
Plasma standards (in untreated rat plasma) included a blank and ranged from
0.05 to 5 mg/ml, with the lower limit of quantification being 0.05 mg/ml. The
accuracy of the assay was determined with replicates of three at 0.05, 0.1, 1, Results
and 10 mg/ml and was 100.1, 97.4, 97.2, and 96.4%, respectively. The Sulfuric Acid Loading. The optimal concentration of doxy loaded
coefficients of variation on the replicates of three at 0.05, 0.1, 1, and 10 mg/ml into the DPPC liposomes was obtained using 4 ml of 8.25 mg/ml doxy
were 11, 6, 16, and 7%, respectively. loaded into liposomes containing 3.0 M sulfuric acid loaded for 24
Liver tissue concentrations of doxy were determined by Liquid Chroma- hours at 22C, with a percentage capture of 97.77%. Doxy loading
tography Mass Spectroscopy, as previously described for plasma. The untreated
using 1 ml of 33 mg/ml produced a preparation with 14.77% capture.
rat liver was fortified with doxy at zero and six concentrations ranging from 0.1
to 50 mg/ml for the standard curve. The tissue standard curve was accepted if Doxy loading approached 100% using 4 ml of doxy solution at
the measured concentration in at least four out of six standards was within 15% a concentration of 8.25 mg/ml, and this method was used for loading
of the actual concentration. Approximately 1 g of tissue was weighed within the liposome preparations used for pharmacokinetics studies in rats.
Loading the same two concentrations for 1 hour and 48 hours at 22 or

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

0.02 g and added to the internal standard solution (500 ng/ml oxytetracycline
final concentration) and homogenized (Polytron PT 10-35 GT; Kinematica, 55C did not improve the milligrams of drug captured (Table 1).
Inc., Bohemia, NY). The homogenates were vortexed and held for 1 hour at Optimal loading conditions for doxy loaded in 3.0 M sulfuric acid–
4C. Two milliliters of 20% trichloroacetic acid solution were added to the containing liposomes with a sphing/cholesterol shell were found to be
homogenate and vortexed, followed by 10 ml of citrate buffer (10 mM, pH 4) 4 ml of 8.25 mg/ml for 24 hours at 22C. The average percentage
and then it was sonicated for 10 minutes, followed by centrifugation at 2500g capture (n = 3) was 43.87.
for 20 minutes. One milliliter of the supernatant then underwent solid phase
Particle Size. The Z-average diameter was 3178 nm.
extraction (SPE) (Strata C18, 3 ml; Phenomenex, Torrance, CA). The SPE was
conditioned with 1.25 ml of methanol, followed by 1.25 ml of deionized water. Thin Layer Chromatography. Analysis of extracted lipids by thin
The sample or standard was loaded, the SPE cartridges were washed with layer chromatography showed the presence of two additional spots,
2.5-ml 5% methanol in deionized water, and 1.25 ml of methanol was used for indicating that significant amounts of breakdown products of DPPC
elution. The eluate was evaporated to dryness under an air stream at 40C. The were present after preparation.
evaporated samples and standards were then reconstituted with 0.2 ml of In Vitro Leakage. Leakage from dialysis bags containing DPPC/
methanol. cholesterol liposomes loaded at 33 mg/ml was faster than for DPPC
Pharmacokinetic Analysis. Pharmacokinetic analysis was performed using LE-doxy loaded using 8.25 mg/ml doxy (Fig. 1A). The DPPC LE-
computer software (WinNonlin 5.2; Pharsight Corporation, Mountain View, doxy made with 8.25 mg/ml leaked less than 45% of the loaded doxy
CA). The parameters determined included area under the curve (AUCall), at the 100-hour time point (Fig. 1A), whereas the DPPC LE-doxy
maximum serum concentration (Cmax), mean residence time (MRTlast), and
made using 33 mg/ml leaked almost 70% of its loaded doxy by the
time to maximum serum concentration (Tmax) (Julious and Debarnot, 2000).
Geometric mean, minimum, median, and maximum were calculated for each same time point. Leakage of DPPC LE-doxy loaded using 4 ml of 8.25
parameter. mg/ml (18.4% in 264 hours) was slower compared with sphing LE-
Histology. The fixed tissue was cut, placed in cassettes, and placed into 10% doxy loaded with the same concentration of doxycycline (21.2% at
neutral buffered formalin until processed. Slides were stained with H&E. The 264 hours) (Fig. 1B). Both LE-doxy in DPPC and sphing would make
pathologist reading the slides was blinded to the group and origin of the tissues. viable preparations to perform initial pharmacokinetics studies in rats.
Cytotoxicity and Bactericidal Properties of LE-Doxy: Cytotoxicity Pharmacokinetics in ACI Rats. After injection, time to peak
Assay. J774A.1 cells were seeded into 24-well culture plates at a density of concentration (Tmax) was 1 hour for DPPC, sphing, and STD-doxy
1.8  105 cells/well. Cells were allowed to adhere to the plates for 16 hours at (Table 2). The MRT of DPPC was highest at 111.78 hours, followed
37C in a humidified 5% CO2 atmosphere. The culture medium, RPMI1640 by sphing (56.00 hours) and STD-doxy (6.86 hours). Serum samples
plus 10% fetal bovine serum, was removed by aspiration, and the cells were
from rats administered sphing LE-doxy had the longest time for the
overlaid with the same medium supplemented with different concentrations of
either STD or DPPC LE-doxy. Treatments were done in triplicate, and medium area under the concentration time curve (120 hours), with DPPC LE-
without the drug was used as a control. Plates were incubated for 24 and 48 doxy at 70.39 hours and STD at 6.36 hours. Serum peak concentration
hours, and cells were harvested using a cell scraper, stained with Trypan blue,
and counted using a hemocytometer. TABLE 1
In Vitro Activity against M. smegmatis. J774A.1 cells (1.8  105/well)
Doxycycline loading using 3.0 M sulfuric acid–loaded 20 mM DPPC/
were allowed to adhere to 24-well plates, as described above for the cholesterol liposomes
cytotoxicity assay. The medium was removed by aspiration, and the cells
were infected with M. smegmatis in RPMI medium + 10% fetal bovine serum Time/Temperature/Input Concentration
Milligrams Captured
(Ghosh et al., 2013) at a multiplicity of infection of 1 bacterium/cell. After 3 of Doxycyclinea
hours, extracellular bacilli were washed away, and STD or DPPC LE-doxy was h/C per mg/ml %
added in five concentrations. The medium without the drug was used as
1/55/33 0.401 (4.86)
a negative control. The efficacy of STD and DPPC LE-doxy was determined by 1/55/8.25 1.54 (74.69)
two methods, colony-forming unit counts were taken at 24 and 48 hours to 24/22/33 1.21 (14.77)
determine the number of viable bacteria and cover slips with adherent, and 24 /22 /8.25 2.02 (97.77)
infected J774A.1 cells were stained with propidium iodide for fluorescence 48/ 22/33 3.199 (38.77)
microscopy counts. Bacteria inside the cells were stained using auramine. The 48/22 /8.25 1.44 (69.57)
total cells and infected cells were counted at 24 and 48 hours. Replicates were a
Loading conditions were 1, 24, or 48 hours on the shaker at either 55 or 22C. A doxycycline
done in triplicate wells. Plates were not counted if microscopic growth was so concentration of either 8.25 or 33 mg/ml was added to the liposome mixture.
 Loading Liposome Doxycycline Pharmacokinetics Efficacy 1239
was 0.72 mg/ml. The plateau concentrations for DPPC and sphing LE-
doxy were $ 0.2 mg/ml.
Liver samples were collected at euthanasia of the rats 4 weeks after
initial injections were performed. The average concentration of STD-
doxy in the liver was 0.093 mg/g (S.D. 6 0.0048). The average
concentration for sphing LE-doxy was 0.174 mg/g (S.D. 6 0.0505)
and that for DPPC LE-doxy was 0.169 mg/g (S.D. 6 0.0850).
Comparisons between hepatic concentrations of doxy in rats
administered STD-doxy and sphing LE-doxy (P = 0.08; Mann-
Whitney Test) and between rats administered STD-doxy and DPPC
LE-doxy (P = 0.086; Mann-Whitney Test) approached, but did not
achieve, statistical significance. Hepatic concentrations of doxy in rats
administered DPPC or sphing LE-doxy were not significantly different
from one another (P = 0.88; Mann-Whitney Test).
Gross Appearance of Injection Sites. A small crust was noted on
one rat injected with STD-doxy, a second rat injected with STD-doxy
had hardened skin and a crust with a superficial slough, and one rat
injected with DPPC LE-doxy had a minor crusted lesion at the
injection site. None of the rats injected with sphing LE-doxy had

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

grossly visible lesions.
Skin Histology. Histopathologic examination was performed on the
injected sites, noninjected skin, and s.c. tissues from the same leg of
all rats in all groups. The pathologist who evaluated the samples (R.S.)
was blinded to the group assignments. Skin samples were classified as
having no lesions, histiocytosis, or miscellaneous inflammation
(Table 3). Rats injected with STD-doxy, DPPC LE-doxy, or sphing
LE-doxy had mild inflammatory lesions in both injected and
noninjected skin samples. There were very few comparisons between
lesions that were statistically significantly different from one another
(Table 3) (Fisher’s exact test). The predominant lesion present in rats
Fig. 1. In vitro leakage of different formulations of LE-doxy. In vitro leakage of administered DDPC LE-doxy, sphing LE-doxy, or blank saline-
liposomes loaded using 33 mg/ml doxy (squares) or 8.25 mg/ml doxy (diamonds) in
DPPC-cholesterol liposomes (A). In vitro leakage comparison between liposomes containing liposomes was a histiocytic infiltrate with foamy cytoplasm
composed of sphing cholesterol (squares) versus DPPC cholesterol (circles) (B). and intensely stained nuclei (Fig. 3, A, C, D, and F). Low numbers
of lymphocytes were sometimes in a few scattered regions or around
vessels in the areas of histiocytosis. One sphing LE-doxy–injected
was highest in sphing LE-doxy (6.59 mg/ml), followed by DPPC LE- animal exhibited moderate lymphocytic infiltrates. The histology of
doxy (2.05 mg/ml) and STD (0.74 mg/ml). Figure 2 is a graphical the skin of the DPPC LE-doxy–injected animal with the grossly noted
representation of the serum concentrations. All formulations were at skin crust exhibited epidermal hyperplasia and adnexal atrophy along
their peak plasma concentration by 1 hour. The steepest declines with connective tissue fibroplasia and granulation tissue formation,
occurred in the first 24 hours, with the slope becoming much slower which is consistent with a healing wound overlying histiocytosis.
for DPPC and sphing LE-doxy after about 72 hours, continuing at There was a wide range of lesion extent, and many injected rats did
roughly the same concentration until the last time point of 336 hours. not have histiocytic lesions (Fig. 3, B and E). The two rats
The DPPC plasma concentration rises slightly over that of sphing LE- administered standard doxy that had grossly visible inflammatory
doxy at 168 hours, maintaining a higher concentration through the last lesions in the first week postinjection did not have any lesions
time point. The peak concentration (Cmax) of DPPC LE-doxy was 2.01 apparent during collection at 4 weeks postinjection (data not shown).
mg/ml. Cmax of sphing LE-doxy was 6.49 mg/ml and that of STD-doxy Cytotoxicity and Bacteriocidal Properties of LE-Doxy: In Vitro
Model of Cell-Associated M. smegmatis Infection. The untreated
TABLE 2 control wells grew to 600,000 J774A.1 cells after 24 hours of incubation
Pharmacokinetics of nonencapsulated STD-doxy and DPPC or sphing LE-doxy (Fig. 4A). There were significantly more live, uninfected cells in the
liposomes in rat serum wells containing STD-doxy compared with DPPC LE-doxy at the 0.05
mg/ml concentration, in the control (0 mg/ml) at 24 hours, and in the
Treatment Parameter Units Geometric Mean Minimum Median Maximum
wells containing STD-doxy compared with DPPC LE-doxy at 0.01 and
DPPC AUCall h*mg/ml 69.34 57.30 66.98 90.28 0.25 mg/ml at 48 hours (Fig. 4, A and B). There were significantly fewer
Sphing AUCall h*mg/ml 118.96 89.88 127.56 132.51 live J774A.1 cells in the wells containing 6.25 mg/ml compared with the
STD AUCall h*mg/ml 6.22 4.81 6.19 8.27
DPPC Cmax mg/ml 2.01 1.42 2.25 2.28 control (0 mg/ml) at 24 hours. All concentrations of STD and DPPC LE-
Sphing Cmax mg/ml 6.49 4.81 6.58 8.01 doxy (0.01, 0.05, 0.25, 1.25, and 6.25 mg/ml) were noncytotoxic to
STD Cmax mg/ml 0.72 0.55 0.76 0.89 uninfected J774A.1 cells (Fig. 4, A and B).
DPPC MRTlast Hours 111.78 89.62 111.53 141.66
Sphing MRTlast Hours 56.00 45.72 54.36 73.80
Concentrations of doxy used in M. smegmatis–infected J774A.1
STD MRTlast Hours 6.86 6.27 6.87 7.51 cells were based on the results of a microplate assay, indicating that
DPPC Tmax Hours 1.00 1.00 1.00 1.00 both STD and DPPC LE-doxy had an Mean Inhibitory Concentration
Sphing Tmax Hours 1.00 1.00 1.00 1.00 (MIC) of 8–16 mg/ml, respectively, for M. smegmatis grown in
STD Tmax Hours 1.00 1.00 1.00 1.00
Middlebrook 7H9 liquid medium (data not shown). Live-infected cells
1240 Franklin et al.

(Fig. 5A). There was a significantly larger percentage of infected cells

in the wells containing control blank liposomes compared with the
0 mg/ml control for STD-doxy at 24 hours. However, there was
a significantly lower percentage of infected cells in the wells
containing DPPC LE-doxy at 2 and 4 mg/ml compared with STD-
doxy at 24 hours (Fig. 5A). There was a higher percentage of infected
cells in wells containing 2, 4, and 16 mg/ml compared with the 0 mg/ml
control at 24 hours (Fig. 5A). There were no surviving cells to count
in the untreated wells (0 mg/ml STD-doxy) or in the wells treated
with 1 mg/ml STD-doxy at 48 hours (Fig. 5B). There was a significantly
larger percentage of infected cells in the wells containing STD-doxy
at 16 mg/ml at 48 hours (Fig. 4B). The MIC 90 for intracellular
M. smegmatis in the J774A.1 cells of both STD and DPPC LE-doxy
was 2 mg/ml (Fig. 5C). The MIC 90 for intracellular M. smegmatis
in this cell line for DPPC LE-doxy at 48 hours was 16 mg/ml (Fig.
5D). The STD-doxy–treated cells reached close to the MIC 50 at
Fig. 2. Serum concentrations and pharmacokinetic data in ACI rats. Rats were 16 mg/ml.
administered STD-doxy at 5 mg/kg (circles), DPPC LE-doxy (squares), or sphing
LE-doxy (triangles) preparations at 50 mg/kg. All formulations were administered

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

s.c. Blood samples were taken before injection for control and from 1 to 336 hours Discussion
(14 days) after injection. Serum was separated, removed, and stored at 220C until
assayed by high-performance liquid chromatography. Sulfuric acid created an effective gradient for movement of the drug
across the liposomes, with less leakage than previous formulations of
liposomal LE-doxy (Sangaré et al., 1998). Loading for 24 hours at
were present in all of the treatments at 24 hours. There were room temperature and at 8.25 mg/ml gave us the optimal capture per
significantly more live cells in the wells containing STD-doxy micromolar of lipid (Table 1). Sulfuric acid–loaded LE-doxy in DPPC
compared with LE-doxy in the wells containing 1, 4, 8, and 16 mg/ml or sphing produced an extended release of at least 2 weeks in vivo in
and control (0 mg/ml) at 24 hours. The number of live-infected cells in rats (Fig. 2; Table 2), and the DPPC formulation had antimicrobial
the wells containing 8 and 16 mg/ml STD-doxy was significantly activity against M. smegmatis at 48 hours (Fig. 5).
lower than the number of cells in the control (0 mg/ml) at 24 hours We developed a novel loading methodology, resulting in liposomes
(Fig. 4C). There were no live cells left in the wells containing 1 mg/ml with higher drug loading and longer retention times than passive
STD-doxy or in the 0 mg/ml control at 48 hours, but the wells aqueous capture (Omri and Ravaoarinoro, 1996; Omri et al., 1995;
containing 1 mg/ml DPPC LE-doxy and blank liposomes contained Sangaré et al., 1998, 1999, 2001a,b). The effectiveness of a gradient-
live cells at that time point (Fig. 4D). There were significantly more loading system is highly dependent on the chemistry of the drug being
live-infected cells in the wells containing 2 mg/ml STD-doxy loaded, its interactions with the liposomal membrane, and the
compared with DPPC LE-doxy, but there were significantly more compound(s) used to form the gradient. Gradient-loading systems in
live-infected cells in the wells containing 8 mg/ml DPPC LE-doxy response to a pH gradient have been developed for ciprofloxacin and
compared with STD-doxy. There were significantly fewer live- other drugs, but the previously described methods for establishing the
infected cells in the wells containing 4, 8, and 16 mg/ml STD-doxy gradient and loading the liposomes differ from the method described
compared with 2 mg/ml STD-doxy (the first concentration for which in the current manuscript (Maurer et al., 1998; Maurer-Spurej et al.,
there were live cells to count) at 48 hours (Fig. 4D). 1999). Our loading and leakage results were better than published LE-
The negative control and concentrations from 1 to 16 mg/ml STD or doxy formulations formed by passive aqueous capture (Sangaré et al.,
DPPC LE-doxy had between 10 and 20% infected cells at 24 hours 1998, 1999, 2001a,b) (Tables 1 and 4). Our loading efficiency is

TABLE 3
Results of histologic evaluation of skin samples obtained at 4 weeks postinjection from ACI rats injected s.c. with STD or LE-doxy

Miscellaneous Inflammation/
Group No Lesions/Number of Samples Histiocytosis/Number of Samples
Number of Samples

% % %
STD-doxy
Injected 4/8 (50)a 0/8 (0) 2/8 (25)
Noninjected 2/8 (25) 0/8 (0) 1/8 (12.5)
DPPC
Injected 0/14 (0)a,b 5/14 (35.7) 2/14 (14.3)
Noninjected 3/14 (21)c 0/14 (0) 5/14 (35.7)
Sphing
Injected 4/12 (33)b 3/12 (25) 0/12 (0)
Noninjected 8/12 (67)c 1/12 (8.3) 0/12 (0)
a
Injected skin samples from rats that received STD-doxy had a greater probability of having no lesions compared with injected samples receiving LE-doxy in DPPC
liposomes (Fisher’s exact test; P = 0.019).
b
Injected skin samples from rats that received LE-doxy in sphing liposomes had a marginally greater probability of having a lesion compared with injected skin
samples from rats that received LE-doxy in DPPC liposomes (Fisher’s exact test; P = 0.066).
c
Noninjected skin samples from rats that received LE-doxy in sphing liposomes had a greater probability of having no lesions when compared with noninjected skin
samples from rats that received LE-doxy in DPPC liposomes (Fisher’s exact test; P = 0.052).
 Loading Liposome Doxycycline Pharmacokinetics Efficacy 1241

Fig. 3. ACI rat injected with sphing LE-doxy liposomes

with mild to moderate histiocytic infiltration of the sub-
cutaneous space (A); rat injected with DPPC LE-doxy with
minimal cellular infiltrates (B); and rat injected with blank
sphing-cholesterol liposomes with moderate cellular infil-
trates (C); 4, bar = 200 mm. ACI rats injected with the

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

same formulations as (A–C) at 20, bar = 50 mm, sphing
LE-doxy liposomes (D), DPPC LE-doxy (E), and blank
sphing-cholesterol liposomes (F). Cellular detail at the
higher magnification includes the foamy cytoplasm of
histiocytes with intensely stained nuclei (D, arrow), and
the accumulation of histiocytes was not limited to any one
group of liposome-injected rats, including those adminis-
tered blank liposomes (F).

10–40 times more efficient than the previously published methods 2010). The inverse of the fraction of hydromorphone loaded versus
(Table 4). Previous formulations of doxy in DPPC/cholesterol hydromorphone added is linear, and the slope is the inverse amount of
liposomes used mannose or glucose solutions at a pH of 6.0 as the amount of ammonium ions present in the liposome (Tu et al.,
nonionic hydrating solutions, and these formulations have encapsu- 2010). The fraction of the drug loaded decreases as the amount of the
lation efficiencies of nearly 10-fold less than the sulfuric acid–loading drug added increases, which was not previously described (Zhigaltsev
technique (Budai et al., 2009). Ciprofloxacin was loaded into DPPC/ et al., 2006). Ammonium sulfate gradient loading without pre-
cholesterol large unilamellar vesicles across a transmembrane pH cipitation produced a concentrated drug within the liposome (Tu et al.,
gradient. The drug forms small stacks in the interior of the liposome 2010). There was no movement of cations across the liposome bilayer.
and precipitates under certain conditions (Maurer et al., 1998; Maurer- Ammonium (NH4) donates a hydrogen ion to the drug, forming
Spurej et al., 1999). Drug loading and precipitation were studied using ammonia (NH3), which diffuses through the lipid bilayer into the
ciprofloxacin and vinorelbine using a hydroxybenzenesulfonate or solution. The unionized drug is brought into the liposome across the
MgSO4 gradient (Zhigaltsev et al., 2006). Drug precipitation inside the gradient.
liposome led to better drug retention (Zhigaltsev et al., 2006). Their We hypothesized doxy would follow this equation, and using
loading equation predicted that the percentage of drug in the liposome hydrogen ions rather than ammonia would further improve the
is not dependent on the initial concentration of the drug when there is loading. We loaded 8.25 mg/ml doxy ( 4 ml for a 33-mg drug) into
no precipitate. (Zhigaltsev et al., 2006). The percentage of released DPPC/cholesterol liposomes at 22C, which was well below the phase
drug is dependent on the initial drug concentration, and equilibrium transition temperature of the lipid (Table 1). The loading efficiency for
between the soluble drug and precipitate follows a zero-order process doxy using 3.0 M sulfuric acid was lower for sphing/cholesterol
if the drug precipitates in the liposome (Zhigaltsev et al., 2006). Tu liposomes compared with DPPC/cholesterol, but it was better than
et al. used an ammonium sulfate gradient–loading system to describe other published results (Tables 1 and 4). Sphing LE-doxy loading
the liposomal loading of the drug based on the amount of ammonium under the same conditions was less efficient, probably due to the
ion in the liposomes and the initial amount of drug added (Tu et al., higher phase transition temperature of the sphing, especially since this
1242 Franklin et al.

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

Fig. 4. Cytotoxicity of doxy in uninfected J774A.1 macrophage cells (A and B) and M. smegmatis–infected J774A.1 macrophage cells (C and D) at 24 (A and C) and 48 (B
and D) hours after culture. Black bars indicate STD-doxy, and gray bars indicate LE-doxy. Significant differences between cells treated with LE or STD-doxy are indicated as
*P # 0.05, **P # 0.01, or ***P # 0.001. Dagger (A, C, and D) indicates a significantly different number of cells between cultures treated with STD-doxy, in which the
numbers of drug-treated cells are less than the zero-concentration control or the lowest drug concentration where cells were present to count (2 mg/ml) (P # 0.05).

sphingolipid was used along with cholesterol to form the lipid described by a number of investigators (Blanchard et al., 1975; Saivin
membrane (Szoka and Papahadjopoulos, 1980). The phase transition and Houin, 1988; Kelly et al., 1992; Sangaré et al., 2001a,b; Selliah
temperature of the DPPC-cholesterol membranes would be expected and Ravaoarinoro, 2004; Rolain et al., 2005; Agwuh and MacGowan,
to be lower than sphing, the membranes would be less rigid, and pH 2006; Zozaya et al., 2013; Gutierrez et al., 2014). We obtained lower
gradient loading would be expected to be faster for a gradient of the values of AUC, Cmax, and MRT for STD-doxy when compared with
same number of pH units between the inner liposome compartments the LE-doxy formulations. Rats administered sphing LE-doxy had
and the outer medium (Szoka and Papahadjopoulos, 1980; Maurer a higher Cmax and AUC compared with those administered DPPC. The
et al., 2001). Loading of sphing liposomes was still adequate for the MRT of DPPC LE-doxy was greater than that for sphing LE-doxy.
preparation to be used in in vitro leakage tests and pharmacokinetics These differences are due to the lipids used for the membranes and the
studies in rats. DPPC LE-doxy liposomes were large, having amount of drug loaded in the liposomes. The higher Cmax for the
a Z-average diameter of 3178 nm, and liposomes made using shaking liposomal formulations was attributed to an initial burst of drug
without other disruptive techniques, such as sonication or extrusion, leakage in both the in vitro and in vivo formulations of LE-doxy and
would be expected to be multilamellar (Szoka and Papahadjopoulos, higher doses in comparison with STD-doxy. The refrigerated
1980). Large, multilamellar liposomes administered subcutaneously liposomal preparations are put into 23C, pH 7.4, saline or
either remain at the injection site and are engulfed by macrophages or approximately 37C, pH 7.4, subcutaneous tissues. The change in
are distributed through lymphatic channels (Oussoren and Storm, pH and temperature caused the drug to efflux until the internal
1999, 2001). Thin layer chromatography results showed the presence liposomal pH was stabilized. The initial burst of leakage tended to
of two additional components, suggesting that DPPC breakdown inhibit further leakage as the internal liposomal pH dropped.
products were present (Supplemental Fig. 1). Given the exposure of The 4-week postinjection liver samples had identifiable levels of
the lipid to acid at elevated temperatures during liposome preparation, doxy. The average concentration for sphing LE-doxy was 0.174 mg/g
it is not unexpected that some lipid would be hydrolyzed in this way (S.D. 6 0.0505) and that for DPPC LE-doxy was 0.169 mg/g (S.D. 6
(Grit et al., 1993a,b). The effect of the presence of these two lipid 0.0850). STD-doxy has a bioavailability of over 80% (Agwuh and
degradation products on the liposome membrane integrity in our MacGowan, 2006). Excretory organs, including the liver, contain the
loading system is presently unknown. The pharmacokinetics (Fig. 2; highest concentrations (Blanchard et al., 1975; Saivin and Houin,
Table 2) of the liposome preparations and liver concentrations at 4 1988). The serum concentration of injectable doxy reaches a maximum
weeks demonstrated that adequate concentrations of intact liposomes 2–4 hours after injection (Blanchard et al., 1975). The liver/serum
were present in the formulations tested. ratio decreases until 8 hours and then rises. These two peaks correlate
Pharmacokinetic analysis of STD-doxy and previous LE-doxy with the low peaks found in the kidney (Blanchard et al., 1975). The
formulations using different loading methodologies has been MIC of many microorganisms is variable, the Cmax of DPPC and
 Loading Liposome Doxycycline Pharmacokinetics Efficacy 1243

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

Fig. 5. Colony-forming units and infected cell counts. Fluorescence microscopy of M. smegmatis–infected J774A.1 cells stained with propidium iodide and auramine stains
at 24 (A) and 48 (B) hours. Percentage viable M. smegmatis (colony-forming units) at increasing concentrations of STD (squares) or LE-doxy (diamonds) at 24 (C) or 48 (D)
hours. Black bars indicate STD-doxy, and gray bars indicate LE-doxy. Dotted lines (C and D) are MIC 50 (gray dots) and MIC 90 (black dots). Significant differences
between cells treated with LE or STD-doxy are indicated as *P # 0.05, **P # 0.01, or ***P # 0.001. Dagger (A) indicates a significantly different percentage of infected
cells between cultures treated with STD-doxy, in which the percentage of drug-treated cells is more than the zero-concentration control (P # 0.05).

sphing (2.01 and 6.49 mg/ml, respectively) in serum was above the and Ravaoarinoro, 2004). Tissues for this study were homogenized in
published MIC of organisms, such as Streptococcus pneumoniaie, 1 ml of sterile water, and the homogenate was evaluated for bacterial
Group A streptococci, and Staphylococcus aureus (Agwuh and inhibition (Selliah and Ravaoarinoro, 2004). Our 4-week postinjection
MacGowan, 2006). The MICs for Coxiella burnetii are reported from liver concentrations were lower but were based on the grams wet
1 to 4 mg/ml, levels which fall within our serum levels (Rolain et al., weight of tissue.
2005). A B-cyclodextrin–based matrix formulation of long-acting STD-doxy has a pH between 1.8 and 3.3 and is not recommended
doxy had a Cmax of 2.8 6 0.3 in dogs (Gutierrez et al., 2014). Mice for i.m. or s.c. use due to the potential for injection site pain, irritation,
infected with Chlamydia trachomatis received intramuscular injec- and tissue necrosis (Zozaya et al., 2013). A nonpainful injection site
tions of cationic liposomal doxy for 30 days using a 10 mg/ml bulge was reported in dogs receiving a long-acting formulation,
solution. The Cmax was 218.75 mg/ml at 48 hours (sera), 18.25 mg/ml although histology was not performed (Gutierrez et al., 2014). There
at 24 hours (liver), and 2.3 mg/ml at 12 hours (genital organs) (Selliah were no skin reactions in our previous studies with liposomal

TABLE 4
Published encapsulation efficiencies of various liposomal-antibiotic formulations

Drug Loading Milligrams

Drug Name Lipid mg/mM Citation
Input Efficiency Loaded

mM mg %
Doxycycline 90 LEC anionic 2 49.17 0.98 0.01 Sangaré et al., 1998,
90 LEC cationic 2 28.68 0.57 0.006 Sangaré
90 LEC neutral 2 21.06 0.42 0.004 Sangaré et al., 1998
Doxycycline 90 LEC anionic 2 49.1 0.98 0.01 Sangaré et al., 1999
90 LEC cationic 2 28.68 0.57 0.006 Sangaré et al., 1999
90 LEC neutral 2 21 0.42 0.004 Sangaré et al., 1999
Doxycycline 90 LEC cationic 2 28.68 0.57 0.006 Sangaré et al., 2001a
Doxycycline 2.73 DPPC, phosphate 0.1354 15.98 (pH 6) 0.02 0.007 Budai et al., 2009, 2009
16.18 (pH 7) 0.02 0.008
16.29 (pH 8) 0.02 0.008
Doxycycline 2.73 DPPC, glucose 0.1354 24.28 (pH 6) 0.03 0.01 Budai et al., 2009
28.31 (pH 7) 0.04 0.01
21.2 (pH 8) 0.03 0.01
Doxycycline 2.73 DPPC mannitol 0.1354 39.47 (pH 6) 0.05 0.02 Budai et al., 2009
30.36 (pH 7) 0.04 0.01
39.47 (pH 8) 0.05 0.02
Doxycycline 20 DPPC 8.25 97.77 8.06 0.403 Current study
Doxycycline 80 sphing 33 43.87 14.47 0.18 Current study

LEC, egg lecithin.

1244 Franklin et al.

morphine, hydromorphone, and oxymorphone tested in rats, dogs, and a lack of cytotoxicity to J774A.1 cells, and had in vitro activity against
nonhuman primates (Krugner-Higby et al., 2003, 2009; Smith et al., M. smegmatis. Our sulfuric acid–loading method provided better
2003, 2013). In the present study, STD-doxy–injected skin samples encapsulation and a longer duration than other passive capture-loading
had a greater probability of having no lesions compared with LE- methods for DPPC and sphing liposomes. Further efficacy research is
DPPC doxy–injected samples. Miscellaneous inflammation was planned for in vivo infectious disease models.
diagnosed in areas of noninjected tissue. Histiocytosis was noted
in the blank, DPPC, and sphing liposomes, but clinically did not Acknowledgments
adversely affect the rats. We found few statistically significant The authors thank the School of Veterinary Medicine vivarium staff for
differences between the groups with respect to skin histology animal care; John Haack and Erin Balay for assistance with the liposomal work;
(Fisher’s exact test), and concluded that the LE-doxy preparations and Matt Warner for determining the liver tissue concentration of doxycycline.
were safe to be given at a dose of 50 mg/kg s.c. in rats.
The sulfuric acid–loaded liposomes were very large when they were Authorship Contributions
sized. The values obtained reflect the expected substantial size of Participated in research design: Heath, Krugner-Higby, Franklin, Marcus,
liposomes made by this method. Large liposomes like these would be Talaat.
Conducted experiments: Franklin, Krugner-Higby, Marcus, Sullivan.
expected to be distributed by circulation in the lymphatics and by
Contributed new reagents or analytic tools: KuKanich.
engulfment by macrophages in the subcutis based on previous Performed data analysis: Franklin, Krugner-Higby, Marcus, KuKanich.
experiments using other liposomal formulations (Oussoren and Storm, Wrote or contributed to the writing of the manuscript: Franklin, Krugner-
1999, 2001). Higby, Heath, Marcus, Talaat, Sullivan, KuKanich.
The efficacy of doxy against cell-associated M. smegmatis at 24

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

hours was similar for both STD and DPPC LE-doxy. However, at 48
References
hours, none of the concentrations of STD-doxy reached the MIC 90,
whereas the liposomal concentration of 16 mg/ml reached the MIC 90, Agrawal AK and Gupta CM (2000) Tuftsin-bearing liposomes in treatment of macrophage-based
infections. Adv Drug Deliv Rev 41:135–146.
indicating a more effective formulation at that time point (Fig. 5, C Agwuh KN and MacGowan A (2006) Pharmacokinetics and pharmacodynamics of the tetracy-
and D). Disparities between MIC values at 48 versus 24 hours may be clines including glycylcyclines. J Antimicrob Chemother 58:256–265.
Blanchard P, Rudhardt M, and Fabre J (1975) Behaviour of doxycycline in the tissues. Che-
attributed in part to loss of intracellular M. smegmatis within dying motherapy 21 (Suppl 1):8–18.
J774A.1 cells. Uninfected J774A.1 cells treated with STD or DPPC Bligh EG and Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J
Biochem Physiol 37:911–917.
LE-doxy continued to grow for 48 hours, and neither treatment was Budai M, Chapela P, Budai L, Wales ME, Petrikovics I, Zimmer A, Gróf P, and Klebovich I
cytotoxic to the cells. There was no clear evidence of dose-dependent (2009) Liposomal oxytetracycline and doxycycline: studies on enhancement of encapsulation
efficiency. Drug Discov Ther 3:13–17.
cytotoxicity in uninfected J774A.1 cells. The concentrations of STD Denis M, Forget A, Pelletier M, Gervais F, and Skamene E (1990) Killing of Mycobacterium
and DPPC LE-doxy formulations used in the experiments in the smegmatis by macrophages from genetically susceptible and resistant mice. J Leukoc Biol 47:
25–30.
current manuscript were all below the cytotoxic threshold (Fig. 4, A Federici TJ (2011) The non-antibiotic properties of tetracyclines: clinical potential in ophthalmic
and B). This experiment was done to determine if the concentrations disease. Pharmacol Res 64:614–623.
Franzblau SG, Witzig RS, McLaughlin JC, Torres P, Madico G, Hernandez A, Degnan MT, Cook
used on infected cells were cytotoxic to the cells and not to find MB, Quenzer VK, and Ferguson RM, et al. (1998) Rapid, low-technology MIC determination
a cytotoxic threshold. There were differences in the numbers of live- with clinical Mycobacterium tuberculosis isolates by using the microplate Alamar Blue assay. J
infected cells at 24 hours that favored STD-doxy (Fig. 4C). Those Clin Microbiol 36:362–366.
Frith JC, Mönkkönen J, Blackburn GM, Russell RGG, and Rogers MJ (1997) Clodronate and
differences were not found at 48 hours (Fig. 4, C and D). Experiments liposome-encapsulated clodronate are metabolized to a toxic ATP analog, adenosine 59-(b,
in which macrophage-origin J774A.1 cells were treated with g-dichloromethylene) triphosphate, by mammalian cells in vitro. J Bone Miner Res 12:
1358–1367.
liposomal clodronate, a liposomal preparation of a drug that kills Ghosh P, Wu CW, and Talaat AM (2013) Key role for the alternative sigma factor, SigH, in the
macrophages, indicated the cytotoxic effects were greater at 48 hours intracellular life of Mycobacterium avium subsp. paratuberculosis during macrophage stress.
Infect Immun 81:2242–2257.
than at 24 hours (Frith et al., 1997). These and the current study’s Grit M, Underberg WJM, and Crommelin DJA (1993a) Hydrolysis of saturated soybean phos-
results may be because the macrophage-origin cells take more than 24 phatidylcholine in aqueous liposome dispersions. J Pharm Sci 82:362–366.
Grit M, Zuidam NJ, Underberg WJM, and Crommelin DJA (1993b) Hydrolysis of partially
hours to phagocytize enough liposomes to have a maximal effect: saturated egg phosphatidylcholine in aqueous liposome dispersions and the effect of cholesterol
cytotoxicity in the case of clodronate and bacterial killing in the case incorporation on hydrolysis kinetics. J Pharm Pharmacol 45:490–495.
Gu Y, Walker C, Ryan ME, Payne JB, and Golub LM (2012) Non-antibacterial tetracycline
of doxy. formulations: clinical applications in dentistry and medicine. J Oral Microbiol 4:19227.
Mycobacterium smegmatis is an environmental mycobacterial Gutiérrez L, Ocampo L, Espinosa F, and Sumano H (2014) Pharmacokinetics of an injectable
long-acting parenteral formulation of doxycycline hyclate in pigs. J Vet Pharmacol Ther 37:
species that has been isolated from humans, often in deep tissue– 83–89.
sequestered infections or in immunocompromised hosts that are Heath TD, Krugner-Higby LA, Smith LJ, Tu S, Kalkhof N, and Franklin RK (2014) inventors,
Comfort Care For Animals, Llc, assignee. Encapsulating liposomes. U.S. patent PCT/US2013/
susceptible to doxy (Wallace et al., 1988). Strains of M. smegmatis 060305. 2014 Mar 27.
isolated from human infections displayed a range of MICs from Henao J, Sánchez D, Muñoz CH, Mejía N, Arias MA, García LF, and Barrera LF (2007) Human
, 0.25 mg/ml to 2–4 mg/ml in non–cell associated assays (Wallace et al., splenic macrophages as a model for in vitro infection with Mycobacterium tuberculosis. Tu-
berculosis (Edinb) 87:509–517.
1988). This is consistent with our study results, which were obtained Hunt CA, Rustum YM, Mayhew E, and Papahadjopoulos D (1979) Retention of cytosine ara-
using the Alamar blue assay after treating M. smegmatis with different binoside in mouse lung following intravenous administration in liposomes of different size.
Drug Metab Dispos 7:124–128.
concentrations of STD or LE DPPC-doxy. Mycobacteria are thought to Julious SA and Debarnot CAM (2000) Why are pharmacokinetic data summarized by arithmetic
persist in the body by establishing stable infections in macrophages means? J Biopharm Stat 10:55–71.
Kelly DJ, Chulay JD, Mikesell P, and Friedlander AM (1992) Serum concentrations of penicillin,
(Wallace et al., 1988). Macrophages phagocytize liposomes well, so doxycycline, and ciprofloxacin during prolonged therapy in rhesus monkeys. J Infect Dis 166:
drugs are targeted to macrophages by liposomal encapsulation (Agrawal 1184–1187.
Krugner-Higby L, KuKanich B, Schmidt B, Heath TD, Brown C, and Smith LJ (2009) Phar-
and Gupta, 2000, Leemans et al., 2001; Sangaré 2001a,b). Our results macokinetics and behavioral effects of an extended-release, liposome-encapsulated preparation
indicate DPPC LE-doxy is effective against cell-associated infection with of oxymorphone in rhesus macaques. J Pharmacol Exp Ther 330:135–141.
Krugner-Higby L, Smith L, Clark M, Heath TD, Dahly E, Schiffman B, Hubbard-VanStelle S,
M. smegmatis for a longer period of time than STD-doxy. Ney D, and Wendland A (2003) Liposome-encapsulated oxymorphone hydrochloride provides
In conclusion, two novel formulations of LE-doxy reached prolonged relief of postsurgical visceral pain in rats. Comp Med 53:270–279.
Kumar R and Malik JK (1998) Some pharmacokinetic parameters and dosage regimens for
measurable and potentially effective concentrations in the serum and a long-acting formulation of oxytetracycline in 6- to 8-month-old male calves. Vet Res Com-
liver for 336 hours, were not damaging to the injection sites, showed mun 22:533–544.
 Loading Liposome Doxycycline Pharmacokinetics Efficacy 1245
Leemans JC, Juffermans NP, Florquin S, van Rooijen N, Vervoordeldonk MJ, Verbon A, van Sangaré L, Morisset R, and Ravaoarinoro M (2001b) [In vitro inhibition of Chlamydia tracho-
Deventer SJH, and van der Poll T (2001) Depletion of alveolar macrophages exerts protective matis growth by liposome-encapsulated cyclines]. Pathol Biol (Paris) 49:53–56.
effects in pulmonary tuberculosis in mice. J Immunol 166:4604–4611. Sapadin AN and Fleischmajer R (2006) Tetracyclines: nonantibiotic properties and their clinical
Maurer N, Fenske DB, and Cullis PR (2001) Developments in liposomal drug delivery systems. implications. J Am Acad Dermatol 54:258–265.
Expert Opin Biol Ther 1:923–947. Selliah S and Ravaoarinoro M (2004) Pharmacokinetics of cationic liposome-encapsulated
Maurer N, Wong KF, Hope MJ, and Cullis PR (1998) Anomalous solubility behavior of the doxycycline in mice challenged with genital infection by Chlamydia trachomatis. Chemo-
antibiotic ciprofloxacin encapsulated in liposomes: a 1H-NMR study. Biochim Biophys Acta therapy 50:17–21.
1374:9–20. Smith LJ, Krugner-Higby L, Clark M, Wendland A, and Heath TD (2003) A single dose of
Maurer-Spurej E, Wong KF, Maurer N, Fenske DB, and Cullis PR (1999) Factors influencing liposome-encapsulated oxymorphone or morphine provides long-term analgesia in an animal
uptake and retention of amino-containing drugs in large unilamellar vesicles exhibiting model of neuropathic pain. Comp Med 53:280–287.
transmembrane pH gradients. Biochim Biophys Acta 1416:1–10. Smith LJ, Kukanich BK, Krugner-Higby LA, Schmidt BH, and Heath TD (2013) Pharmacoki-
Nelson ML and Levy SB (2011) The history of the tetracyclines. Ann N Y Acad Sci 1241: netics of ammonium sulfate gradient loaded liposome-encapsulated oxymorphone and
17–32. hydromorphone in healthy dogs. Vet Anaesth Analg 40:537–545.
Omri A and Ravaoarinoro M (1996) Preparation, properties and the effects of amikacin, Stegemann MR, Sherington J, and Blanchflower S (2006a) Pharmacokinetics and pharmacody-
netilmicin and tobramycin in free and liposomal formulations on Gram-negative and namics of cefovecin in dogs. J Vet Pharmacol Ther 29:501–511.
Gram-positive bacteria. Int J Antimicrob Agents 7:9–14. Stegemann MR, Sherington J, Coati N, Brown SA, and Blanchflower S (2006b) Pharmacoki-
Omri A, Ravaoarinoro M, and Poisson M (1995) Incorporation, release and in-vitro antibacterial netics of cefovecin in cats. J Vet Pharmacol Ther 29:513–524.
activity of liposomal aminoglycosides against Pseudomonas aeruginosa. J Antimicrob Che- Szoka F, Jr and Papahadjopoulos D (1980) Comparative properties and methods of preparation of
mother 36:631–639. lipid vesicles (liposomes). Annu Rev Biophys Bioeng 9:467–508.
Oussoren C and Storm G (1999) Role of macrophages in the localisation of liposomes in lymph Tu S, McGinnis T, Krugner-Higby L, and Heath TD (2010) A mathematical relationship for
nodes after subcutaneous administration. Int J Pharm 183:37–41. hydromorphone loading into liposomes with trans-membrane ammonium sulfate gradients. J
Oussoren C and Storm G (2001) Liposomes to target the lymphatics by subcutaneous adminis- Pharm Sci 99:2672–2680.
tration. Adv Drug Deliv Rev 50:143–156. Wallace RJ, Jr, Nash DR, Tsukamura M, Blacklock ZM, and Silcox VA (1988) Human disease
Raabe BM, Lovaglio J, Grover GS, Brown SA, Boucher JF, Yuan Y, Civil JR, Gillhouse KA, due to Mycobacterium smegmatis. J Infect Dis 158:52–59.
Stubbs MN, and Hoggatt AF, et al. (2011) Pharmacokinetics of cefovecin in cynomolgus Zeidner NS, Massung RF, Dolan MC, Dadey E, Gabitzsch E, Dietrich G, and Levin ML (2008) A
macaques (Macaca fascicularis), olive baboons (Papio anubis), and rhesus macaques (Macaca sustained-release formulation of doxycycline hyclate (Atridox) prevents simultaneous infection
mulatta). J Am Assoc Lab Anim Sci 50:389–395. of Anaplasma phagocytophilum and Borrelia burgdorferi transmitted by tick bite. J Med

Downloaded from dmd.aspetjournals.org at ASPET Journals on July 1, 2015

Rolain JM, Boulos A, Mallet MN, and Raoult D (2005) Correlation between ratio of serum Microbiol 57:463–468.
doxycycline concentration to MIC and rapid decline of antibody levels during treatment of Q Zhigaltsev IV, Maurer N, Edwards K, Karlsson G, and Cullis PR (2006) Formation of drug-
fever endocarditis. Antimicrob Agents Chemother 49:2673–2676. arylsulfonate complexes inside liposomes: a novel approach to improve drug retention. J
Saivin S and Houin G (1988) Clinical pharmacokinetics of doxycycline and minocycline. Clin Control Release 110:378–386.
Pharmacokinet 15:355–366. Zozaya H, Gutierrez L, Bernad MJ, and Sumano H (2013) Pharmacokinetics of a peroral single
Sangaré L, Morisset R, Gaboury L, and Ravaoarinoro M (2001a) Effects of cationic liposome- dose of two long-acting formulations and an aqueous formulation of doxycycline hyclate in
encapsulated doxycycline on experimental Chlamydia trachomatis genital infection in mice. J horses. Acta Vet Scand 55:21.
Antimicrob Chemother 47:323–331.
Sangaré L, Morisset R, Omri A, and Ravaoarinoro M (1998) Incorporation rates, stabilities,
cytotoxicities and release of liposomal tetracycline and doxycycline in human serum. J Anti-
Address correspondence to: Rebekah K. Franklin, 396 Enzyme Institute, 1710
microb Chemother 42:831–834.
Sangaré L, Morisset R, and Ravaoarinoro M (1999) In-vitro anti-chlamydial activities of free and University Ave., Madison, WI 53726. E-mail: [email protected]
liposomal tetracycline and doxycycline. J Med Microbiol 48:689–693.