Final Report
Final Report
Final Report
Due to the Covid-19 pandemic, the Summer Science Program (SSP) was conducted online in the Summer of 2020.
Thus, all data used in this report were collected by participants who participated in previous years of the SSP
Biochemistry program.
1
Abstract
Fungal pathogens are a major threat to global food security, and Verticillium longisporum is one fungal
pathogen that causes immense crop devastation worldwide. The Cdc14 enzyme is an attractive target for
the development of V. longisporum pesticides and novel antifungal drugs in general because of its highly
conserved function, absence in higher plants, and high substrate specificity. The goal of this research is to
design an inhibitor for Cdc14 in Verticillium longisporum. This was done by expressing the target
protein’s gene in E. coli, purifying it using nickel affinity chromatography, and ensuring the protein
exhibited phosphatase activity using computational analysis and experimental assays. Using MOE, a
well-studied Cdc14 enzyme from yeast was used to build a homology model, and the structural similarity
of the active site and overall enzyme prompted comparison of kinetic parameters and substrate specificity.
Due to the conservation of structural elements, similarity of catalytic efficiency, and related specificity,
inhibitor docking studies were conducted based on known yeast substrates. From this, the best inhibitor
was optimized through addition of functional groups to increase its affinity with the enzyme. This
inhibitor can serve as the basis for the development of a novel fungicide that can help alleviate world
hunger.
2
Introduction
Severe malnutrition affects over 820 million people world-wide, and for the past five years, world
hunger levels have been on the rise (1). Over a quarter of the world’s population suffers from moderate to
dire food insecurity, and the basis of this problem lies in a food production barrier: plant disease (1, 2).
Each year, approximately 10 to 16 percent of the global harvest is lost to devastating crop infections, and
from an economic standpoint, these losses total to over 220 billion dollars in the U.S. (2). There are
several causes of crop disease including viruses, pests, and weeds, but of these, fungi and oomycetes
cause the most crop devastation (2, 3). Fungal plant disease disproportionately impacts developing
countries where one-third of food crops are destroyed by these pathogens (4).
One major fungal crop pathogen that targets cruciferous vegetables such as broccoli, cauliflower,
and cabbage is Verticillium longisporum, hereby referred to as V. longisporum (5). Control of V
.
longisporum d isease is particularly costly because it can remain dormant in soil for over 10 years before
germinating in the presence of a host (6). The disease caused by this fungal crop pathogen is known as
Verticillium wilt, and currently, no effective fungicides have been created to control the spread of V.
longisporum because of the “silent” nature of this disease (7). Characterization of disease symptoms and
attribution of these symptoms to V. longisporum has been relatively recent, and we still do not have a
flavus from prior literature is the Cdc14 family (8, 9). The Cdc14 family is broadly categorized under
dual-specificity protein phosphatases, a type of phosphatases that remove a phosphate group from
Cdc14 phosphatases present a promising target for the development of novel antifungal drugs for
a variety of reasons. First, Cdc14’s structure is highly conserved throughout multiple fungal species (11).
3
Previous studies have shown that all Cdc14 phosphatases share a conserved N-terminal core of about 350
amino acids. Also, analysis of Cdc14 crystal structure revealed two conserved domains: domain A that
contributes to substrate specificity and domain B that is responsible for catalytic activity and encompasses
the active site sequence VHCSXGXGR[T/S]G (called PTP signature motif) (11, 12). The high
conservation indicates a possibility that the substrates of this enzyme in each are similar, and that a
general fungicide targeting multiple plant pathogens can be developed (11). The biological function,
biochemical mechanism, and atomic structure of VlCdc14 (the Cdc14 protein found in V. longisporum)
can also be predicted based on data from related fungi such as budding yeast which can be useful in the
characterization of this enzyme and the creation of new fungicides. However, while Cdc14 enzymes are
highly-conserved across fungi and plants, they are absent from higher plants (8). Thus, plants are likely to
Cdc14 sacrificed the structural integrity of the fungal genome (13). Genome instability results in a variety
of diseases, so this provides indication that Cdc14 contributes to V. longisporum’s pathogenic properties
in this manner and is therefore a promising target for antifungal pesticides (14). Lastly, Cdc14 enzymes
are highly specific and preferentially dephosphorylate particular substrates, so designing a specific
hypothesized to target sites following the pattern “phosphoserine – proline – X – lysine/arginine” where X
represents any amino acid (15). Specificity is an important characteristic of optimal drug targets because
other enzymes in plants and animals will likely not be affected by the fungicide.
The purpose of this research is to create a structural model of Cdc14 in V. longisporum, calculate
the kinetic parameters of this enzyme, and evaluate its substrate specificity to design a small-molecule
competitive inhibitor. By building on knowledge of biological function and biochemical properties from
4
prior literature, this work provides the first step in designing a novel fungicide for V. longisporum,
eventually helping to alleviate the wide-spread effects of world hunger and food insecurity.
Methods
From the results of a genetic screen of V. longisporum, the gene sequence of an unknown protein
was obtained. A GenScan and subsequent protein BLAST search was run with the gene sequence to
identify the coding sequence for our target protein. Then, another protein BLAST search was then run to
find similar sequences, and based on a multisequence alignment between our target protein and
well-studied orthologs from organisms such as Saccharomyces cerevisiae, the protein was hypothesized
to be within the Cdc14 phosphatase family. In order to test this hypothesis and study the enzyme further,
the protein was expressed in E. coli bacteria. The pET15b plasmid vector was used, and two directional
cloning restriction enzyme sites were chosen from the plasmid’s multiple cloning sequence (MCS) to use
for the cloning experiment. The chosen restriction enzymes were NdeI and BamHI due to their location
on pET15b and their compatible reaction conditions. After designing forward and reverse primers based
on the protein coding sequence and restriction sites, a map of the recombinant plasmid was created in the
visualization program SnapGene. Considerations such as distance between the N-terminus of the target
Once the V. longisporum g ene was successfully placed in frame with the T7 promoter and lac
operon as validated by SnapGene, the purified VlCdc14 gene was inserted into the pET15b plasmid. The
recombinant plasmid was then used to transform competent E. coli c ells. After the protein was expressed,
the protein was purified using nickel-affinity chromatography. A small sample of the extract is saved for
5
future gel analysis. Only 6xHis-tagged proteins will adhere to the column, and the unbound proteins are
In order to assess the purity of the protein, an SDS-PAGE gel was run. After setting up the gel
with a resolving buffer and a stacking buffer, the first lane of the gel was run with an uninduced whole
cell extract. The subsequent lanes were induced whole cell extract, soluble protein extract, nickel column
flow-through, and pooled nickel elution peak. Using the Image Lab program, the gel electrophoresis
results, and the MW standards we found the molecular weight of the target protein. We compared the
average MW across lanes 6, 7, and 8 with the result from ExPASy, which was obtained by inputting the
Then, a Bradford assay was used to calculate the concentration of our purified protein. A
spectrophotometer was used to measure the absorbance of known amounts of bovine serum albumin
(BSA) at 595nm. From this data, a BSA standard curve was generated and linear regression analysis was
performed. Next, the absorbance of the purified target protein at 595 nm was measured at two different
dilutions: 10 ul of 1:10 dilution and 1:50 dilution. By repeating each dilution 3 times and averaging the
results, the absorbance value was converted to protein concentration based on the BSA standard curve.
After ensuring that the correct protein had been purified and a sufficiently concentrated solution
had been created, a phosphatase assay was utilized to ensure that our target protein had phosphatase
activity. The absorbance at 405 nm of reaction wells with our target protein and pNPP both with and
without a sodium tungstate inhibitor were run, and using a standard curve, these absorbances were
converted into pNP concentration. The rates of pNP formation with and without the inhibitor were
6
After confirming the target protein’s function, a steady-state kinetic assay was performed to
determine the kcat and Km of VlCdc14 with the pNPP substrate. Each reaction well contained 2 μM of
VlCdc14 and a different concentration of pNPP (0, 1, 4, 8, 10, 20, 40, 60, 80 mM). Final absorbance at
405 nm of each reaction well was measured after 15 minutes, and blanks were used to correct for
background signal. A standard curve was utilized to convert absorbance data into pNP concentration, and
dividing pNP concentration by the reaction time gave us an approximation of the initial velocity for each
substrate concentration. The Solver function on Google Sheets was utilized to fit this data to the
Michaelis-Menten equation which returned the kcat and Km values for this particular reaction and
enzyme. Averaging these results across 3 trials returned a prediction for the catalytic efficiency of the
enzyme. These catalytic parameters were necessary to determine reaction conditions for subsequent
assays.
The next step was to determine the target protein’s specificity. For this, identification of the best
predicted substrates for VlCdc14 was needed. However, because the structure of VlCdc14 has not been
Based on the BLAST search done previously, ScCdc14 (the Cdc14 protein in S. cerevisiae) and
VlCdc14 were shown to have a very low E-value, providing evidence that the two proteins are homologs.
Hence, a homology model of VlCdc14 using ScCdc14 as the template was created. After aligning and
superposing the VlCdc14 sequence and ScCdc14 sequence, a color scheme based on RMSD values was
applied to the model, with greener colors representing lower RMSD values and redder colors representing
higher ones. The majority of the model was green, indicating that ScCdc14 and VlCdc14 were similar in
structure.
7
The substrate binding pockets of our homology model and ScCdc14 were then compared in MOE
and shown to be nearly identical. So, VlCdc14 and ScCdc14 were hypothesized to have similar substrate
specificity. To test this, the affinity of VlCdc14 for 24 different phosphopeptide sequences was predicted
based on literature about the specificity of ScCdc14. Then, each of these substrates were tested in a
specificity assay with VlCdc14. 24 reaction wells were used, with 0.5 μM of VlCdc14, 0.1 mM of one of
the phosphopeptides, and 0.8 M buffer. After 10 minutes, malachite green dye was added and the
absorbance at 640 nm was measured. Three trials were conducted for each substrate, and blanks without
enzymes were used to subtract background signals from absorbances. Then, utilizing a standard curve of
the absorbance at 605 nm versus phosphate concentration created using given spectrophotometric data,
absorbances were converted into final concentration of phosphate. Dividing this concentration by the
reaction time returned an approximation of the initial reaction velocity for each substrate with VlCdc14.
These results were used to determine whether or not ScCdc14 and VlCdc14 have similar substrate
specificity.
An inhibitor screen of VlCdc14 with 11 inhibitors that have shown affinity to ScCdc14 was run.
Each experimental reaction well contained 15 mM pNPP, 2 μM VlCdc14, 50 μM of one of the inhibitors
(unless the inhibitor was H7 or I9, then there was 1000 μM), and 0.795M buffer. A positive control with
15 mM pNPP, 2 μM VlCdc14, and 0.8M buffer, and a negative control of just 15 mM pNPP were also
run.
After 15 min, the absorbance at 405 nm of each of these 13 wells was measured. Blanks were also
utilized to subtract background signals from absorbance data. Three trials were done for each reaction
well. After confirming the z-value for the assay was above 0.5, the average percent inhibition for each of
the trials was calculated by comparing each inhibitor’s average absorbance to that of the positive control.
Computational docking of all inhibitors to VlCdc14 was also done in MOE, using our
previously-built homology model. From the percent inhibition value and docking results, the top inhibitor
8
was chosen. To optimize this inhibitor, the 3D electrostatic surface of its interactions was examined in
MOE, and functional groups were in different areas of the inhibitor, based on whether the surface was
hydrophobic, H-bond donors, or H-bond acceptor in that area. After this, MOE was used to predict the
Results
The target gene obtained from a V. longisporum genetic screen was analyzed using a BLAST
search in order to predict the likely function of the protein. As shown in Table 1, the orthologs of the
coding sequence of our target protein are all part of the CDC14 phosphatase family. This suggested that
The sequences from the BLAST search were then used to conduct a multi-sequence alignment to
further support our hypothesis that our target sequence coded for a Cdc14 enzyme. As shown in Figure 1,
a sequence of 13 residues that is highly-conserved in Cdc14 enzymes from various other species was
present in our target gene. Additionally, motifs such as cysteine-lysine and arginine-tyrosine which are
important in the Cdc14 mechanism were conserved (16). This provides evidence that the sequence being
To test whether or not our protein is a Cdc14 phosphatase, we incorporated the gene in a
recombinant plasmid and expressed it in E. coli, as seen in Figure 2. With this, we could create large
amounts of our target protein. After expressing the recombinant plasmid in E.coli and purifying the
protein through nickel affinity chromatography, we ran SDS-PAGE to ensure that our protein was
purified, and these results are shown in Figure 3. Based on the molecular weight markers shown vertically
on the left, the molecular weight of the purified protein was 54.908 kDa. The theoretical molecular weight
from ExPASy was 54.907 kDa, so the percent error is ~0%, suggesting that we purified the correct
protein. This was further supported by the fact that there was only one significant band in lanes 6-8, and
9
the intensity of that band was high. Then, using a Bradford assay, we calculated the purified protein
concentration, and this data is shown in Table 2. Using the average concentration value of 5.255 mg/mL
and the molar mass from the SDS-PAGE of 54.908 g/mol, we calculated our protein concentration to be
97.71 μM.
After ensuring that the correct protein had been purified and a sufficiently concentrated solution
had been created, we used a phosphatase assay to confirm the phosphatase activity of our target protein as
shown in Figure 4. As shown in Table 3, the average specific activity without the sodium tungstate
inhibitor (a known phosphatase inhibitor) was 13.39 uM pNP per min per uM Cdc14 while the average
specific activity with the inhibitor was 3.92 uM pNP per min per uM Cdc14. From this result, we
concluded that the average percent activity inhibited by sodium tungstate was 70.36%. Additionally, we
ran a T-test to measure significance of the difference in reaction velocity between the uninhibited and
inhibited groups, and our T-value of 10.554 was greater than the critical value at a significance level of
0.02. The p-value was 0.0044 (less than 0.01) which indicates that there was a significant difference in
activity between the inhibited and original enzyme. Overall, this confirmed our hypothesis that the target
protein in V. longisporum (which we will hereby refer to as VlCdc14) displays phosphatase activity.
Next, a steady-state kinetic assay was performed, and from the Michaelis-Menten curve shown in
Figure 5, we calculated the kcat value to be 0.0962 s-1, Km to be 15.5 mM, and vmax to be 0.192uM/sec. The
catalytic efficiency value was used to determine the conditions for the subsequent specificity assay. To
examine our protein’s substrate specificity, we began with probable substrates from S. cerevisiae. This is
because as stated previously, from a multi-sequence alignment, VlCdc14 and ScCdc14 (the Cdc14 protein
found in S. cerevisiae) were shown to have a low E-value, suggesting that they are homologs.
Additionally, the optimal RMSD values throughout most of the structure shown in the homology model
(Figure 6a), indicates that VlCdc14 and ScCdc14 are very similar in structure. As shown in Figure 6b, the
phi-psi plot shows very few outlier residues. We then aligned and superposed the active site specifically
10
as shown in Figure 6c, and found that there was little difference between the binding pockets. The sample
phosphopeptide (Serine-Proline) that binds well to ScCdc14 was predicted to bind well to VlCdc14 in
MOE. From this, we hypothesized that VlCdc14 and ScCdc14 have similar substrate specificity.
To test this hypothesis that VlCdc14 and ScCdc14 have similar substrate specificity, we evaluated
the activity of our enzyme with 24 different substrates. As shown in Table 4, the first substrate was the
Yen1 peptide which is a known ScCdc14 substrate containing motifs such as a phosphorylated serine and
several basic residues. The Yen1 substrate showed a high average rate of reaction in VlCdc14, indicating
that our target protein has similar substrate specificity to ScCdc14. The subsequent 23 substrates
contained modifications to the motifs present in Yen1 in order to identify the patterns important to
VlCdc14 binding. Based on the results shown in Figure 7, substrates 7, 9, 14, 21, and 24 performed well,
and all of them followed the pattern “{pS}-proline-x-lysine” where x is any amino acid (Table 4 shows
Based on the desirable motifs identified in the specificity assay, 12 inhibitors of ScCdc14 were
identified and tested for effectiveness in an inhibitor screen. Our inhibitor screening assay had a Z-factor
of 0.75 based on positive and negative control data, indicating that the assay was effective enough to
correctly show average percent inhibition. As shown in Table 5, the best inhibitors in terms of average
After experimentally finding our top inhibitors, we used the MOE program to dock several poses
of each inhibitor in the active site. A majority of the G6 inhibitor poses consistently docked within the
VlCdc14 active site (Figure 8A), and the best pose of the G6 inhibitor had a low S-value of -5.27 and it
had an affinity score of -6.53 kcal/mol with VlCdc14. Additionally, as shown in the original ligand
interaction map (Figure 8B), the inhibitor’s SO3 group interacts with Arg341, a residue that is important
in Cdc14 enzymatic mechanism (16). However, no interaction is formed with Cys335, another residue
11
that launches a nucleophilic attack in the Cdc14 mechanism. We optimized the ligand by adding two
amine groups, one fluorine and one carbon atom, an imidazole group, and a hydroxyl group which then
formed contacts with Cys335 and several other surrounding residues in the binding pocket. These
modifications improved the affinity score to -9.58 kcal/mol. The interactions between the optimized
Discussions
The structure and specificity Cdc14 phosphatases have previously been characterized in the
fungal species S. cerevisiae. These prior results show that arginine and cysteine residues in the active site
are crucial to the biochemical mechanism of Cdc14, and that Cdc14 phosphatases preferentially bind to
substrates containing the serine-proline-x-lysine binding motif where x represents any amino acid. Our
work expands these conclusions to VlCdc14 by showing the similarity in structure between VlCdc14 and
ScCdc14, demonstrating VlCdc14’s specificity and proposing a novel competitive inhibitor design to
create a fungicide.
We began by hypothesizing that the target sequence obtained from the V. longisporum genetic
screen was likely a Cdc14 enzyme. Analysis of BLAST results and multi-sequence alignments served as
the basis for this hypothesis because important motifs were conserved. The phosphatase activity assay
and inhibition by tungstate confirmed that our target protein was indeed a Cdc14 enzyme that hydrolyzes
After confirmation of phosphatase activity, kcat and Km values were calculated using a steady-state
assay. Compared to the ScCdc14’s kcat value of 4.5 s-1 with the pNPP substrate, our enzyme’s kcat value of
0.1 s-1 was somewhat lower, but our calculated Km value of 15.6 mM was relatively similar to the
accepted value of 10.6 mM. The difference in Km may be explained by the unusually high enzyme
12
concentration of 2 μM used in the steady-state assay which may have resulted in the enzyme not being
completely saturated. The difference in the kcat values could have been due to the experimental
conditions. Additionally, the comparison kcat value is from ScCdc14, so a few differences in structural
characteristics between VlCdc14 and ScCdc14 may have contributed to the difference in kcat.
Next, we needed to determine the substrate specificity of our enzyme. The specificity of our
enzyme is an important assumption behind the design of selective inhibitors. From our homology model,
we hypothesized that ScCdc14 and VlCdc14 have similar substrate specificity. We were able to confirm
our hypothesis with our substrate specificity experiment, which showed that the known ScCdc14 substrate
Yen1 showed a high rate of reaction for VlCdc14. With this, we were able to perform an inhibitor screen
on VlCdc14 using known inhibitors for ScCdc14. Computational docking studies on the top inhibitors
from the screen showed that G6 (containing two benzene rings and one sulfur trioxide group) was our
best. The sulfur trioxide group adopted a similar position as the phosphate group from the original
phosphopeptide substrate in the binding pocket which provides further evidence that G6 is a good
inhibitor. We then made modifications to this inhibitor, increasing interactions between important
residues such as arginine and cysteine. First, an imidazole group was added to mimic the side chain of
proline, a known Cdc14 substrate recognition feature. Next, two amine groups were added to mimic the
original phosphopeptide, and a fluorine atom was added to improve binding orientation (due to fluorine’s
electronegativity). Finally, we added a hydroxyl group to the imidazole ring to mimic the serine
side-chain, a known substrate recognition feature. While these ligand interaction maps and the
improvement in affinity score provide evidence that our modifications create a more effective inhibitor,
further lab testing is required. Future work should test several modified G6 inhibitors for percent
inhibition. A dose response experiment is essential to determine the concentration at which these
13
The work done in this study including creation of a structural model of VlCdc14, kinetic
inhibitor for V. longisporum. By providing evidence to show the highly-conserved nature of the Cdc14
active site, we built upon prior literature to validate that Cdc14 is an effective fungal pesticide target.
Overall, this work can serve as a basis in designing a novel fungicide to target Verticillium longisporum.
Additionally, Cdc14’s conserved structure and substrate specificity may allow the methods we used to be
utilized for the creation of pesticides for other fungal plant pathogens and crop diseases.
References
1. Wise, Timothy. (2019) World Hunger Is on the Rise Chain Reaction. 136: 42–43; doi: 1 0.2301/652732
2. Chakraborty, S., and A. C. Newton. (2011) Climate Change, Plant Diseases and Food Security: an
Overview Plant Pathology. 160: 2-14; doi: 10.1111/j.1365-3059.2010.02411.x
3. Fones, Helen N., et al. (2020) Threats to Global Food Security from Emerging Fungal and Oomycete
Crop Pathogens Nature Food. 6: 332-342; doi: 10.1038/s43016-020-0075-0
4. Almeida, Fausto, et al. (2019) The Still Underestimated Problem of Fungal Diseases Worldwide
Frontiers in Microbiology. 10; doi: 10.3389/fmicb.2019.00214
5. Goss, Erica M., et al. (2011) The Plant Pathogen Phytophthora Andina Emerged via Hybridization of
an Unknown Phytophthora Species and the Irish Potato Famine Pathogen, P. Infestans PLoS ONE. 6; doi:
10.1371/journal.pone.0024543
6. Inderbitzin, Patrik, et al. (2011) Phylogenetics and Taxonomy of the Fungal Vascular Wilt Pathogen
Verticillium, with the Descriptions of Five New Species. PLoS ONE, 6; doi:
10.1371/journal.pone.0028341
14
9. Yang, Guang, et al. (2018) The Aspergillus Flavus Phosphatase CDC14 Regulates Development,
Aflatoxin Biosynthesis and Pathogenicity Frontiers in Cellular and Infection Microbiology. 8; doi:
10.3389/fcimb.2018.00141
10. Gray, C. H. (2003) The Structure of the Cell Cycle Protein Cdc14 Reveals a Proline-Directed Protein
Phosphatase The EMBO Journal. 22: 3524-3535; doi: 10.1093/emboj/cdg348
11. Mocciaro, A., and Schiebel, E.(2010) Cdc14: a highly conserved family of phosphatases with
non-conserved functions? Journal of Cell Science. 123: 2867-2876; doi: 10.1242/jcs.074815
12. Andersen, J N, et al. (2001) Structural and Evolutionary Relationships among Protein Tyrosine
Phosphatase Domains Molecular and Cellular Biology, American Society for Microbiology. 21:
7117-7136; doi: 10.1128/MCB.21.21.7117-7136.2001.
13. Quevedo, Oliver, et al. (2015) The Transient Inactivation of the Master Cell Cycle Phosphatase
Cdc14 Causes Genomic Instability in Diploid Cells of Saccharomyces Cerevisiae Genetics. 200: 775-769;
doi:10.1534/genetics.115.177626
14. Eissler, Christie et al. (2014) The Cdk/Cdc14 Module Controls Activation of the Yen1 Holliday
Junction Resolvase to Promote Genome Stability Molecular Cell. 54: 80-93, doi:
10.1016/j.molcel.2014.02.012
15. Kobayashi, Junya, and Yoshiyuki Matsuura. (2017) Structure and Dimerization of the Catalytic
Domain of the Protein Phosphatase Cdc14p, a Key Regulator of Mitotic Exit in Saccharomyces
Cerevisiae. Protein Science, 26: 2105–2112; doi: 10.1002/pro.3244
16. DeMarco, Andrew G., et al. (2020) Conservation of Cdc14 Phosphatase Specificity in Plant Fungal
Pathogens: Implications for Antifungal Development. Nature News; doi: 10.1038/s41598-020-68921-3
15
Figures
Table 1. BLAST results show that sequences that are most similar to the target sequence are Cdc14
phosphatases. E-values from sequences across diverse species indicate that our target protein is likely a
Cdc14 phosphatase.
16
Figure 1. Multi-sequence alignment of BLAST sequences shows conserved domain consisting of
cysteine-lysine motif and arginine-tyrosine motif. Alignment of 5 known Cdc14 sequences from
different species with the target V. longisporum s equence shows a highly-conserved domain.
17
Figure 2. Recombinant plasmid map contains V. longisporum t arget gene, 6xHis tag, and lac
operator. The total length of the plasmid was 6457 bp, which is the sum of the original pET15b length
and the target sequence length. To confirm that the recombination was successful, we would need to
perform an analytical digest with restriction enzymes BamHI and NdeI.
18
Figure 3. SDS-PAGE analysis shows successful purification of correct target protein
Lane 1 shows the molecular weight standards, lane 2 shows uninduced whole cell extract, lane 3 shows
induced whole cell extract, lane 4 shows soluble protein extract, lane 5 shows nickel column
flow-through, and lanes 6 through 8 (right-most) show the pure protein pool.
19
Absorbance (Abs)
Amount of Concentration
Trial 1 Trial 2 Trial 3 Average Standard protein (mg/mL)
deviation (mg)
Average 5.255
Table 2. Bradford Assay: protein concentration was 5.255mg/mL.
(a) Absorbance at 595nm of 10uL of purified target protein at 1:10 dilution. The data shows scaled
absorbance (background value subtracted from absorbance value) for 3 trials, the average absorbance,
standard deviation, amount of protein in mg from the BSA standard curve, and the concentration of
protein in mg/mL.
(b) Absorbance at 595nm of 10uL of purified target protein at 1:50 dilution. Same information as (a)
provided.
20
Figure 4. Sodium tungstate inhibitor significantly decreases pNPP hydrolysis rate.
(a). pNP concentration without tungstate inhibitor vs. time. The data shown is the average pNP
concentration value of 3 trials as 30mM pNPP reacted with 0.33μMVlCdc14 enzyme concentration for 30
minutes to form product pNP
(b). pNP concentration with 2 mM tungstate inhibitor, and all other reaction conditions from part (a)
constant.
21
Figure 5. Michaelis-Menten Curve with the Substrate pNPP
With three trials for each substrate concentration in our steady-state kinetics assay, we determined the
average initial reaction velocity for each concentration . This data was then fitted to the Michaelis-Menten
equation: y = ((vmax)*x)/(Km + x). Standard deviation error bars were then applied to the graph. vmax was
calculated to be 0.19 uM/sec, Km was calculated to be 15.6 mM, and kcat was calculated to be 0.1 s-1.
22
Figure 6a. Homology model of VlCdc14 based on a template from S. cerevisiae. (a) T his image from
MOE is colored by RMSD values, where green repeats low RMSD values and white and red represents
higher RMSD values. (b). Ramachandran plot of our homology model shows very few outliers of
phi-psi angles. Seven residues are shown as outliers out of over 400 aligned residues, indicating that
VlCdc14 and ScCdc14 are highly similar in structure. (c) Known ScCdc14 peptide is predicted to bind
well to the VlCdc14 homology model. This indicates that VlCdc14 and ScCdc14 have similar active
sites and bind well to similar substrates.
23
Figure 7. VlCdc14 shows higher specificity to peptides containing phosphoserine, basic residues,
and two adjacent lysine residues. Data shows the average rate of reaction for each substrate. The error
bars indicate the standard deviations among the 3 trials. The sequences of the 24 peptides used is shown
below in Table 3.
Table 4. Naming of 24 substrates tested in specificity assay. {p} represents a phosphorylated amino
acid. Peptide 1 is a known substrate of ScCdc14 called Yen1.
24
Inhibitor ID Average % Inhibition Standard Deviation
A9 73.19004278 11.18072971
D3 59.39751067 5.460849799
D7 9.473698995 5.152410204
E1 29.24552011 2.384421953
G5 28.21225242 35.11390872
G6 75.57955658 7.358674826
G7 93.83538753 5.99806634
H7 75.57976431 7.093802647
I1 72.90038899 5.499119207
I2 37.098296 4.924507476
I9 60.83616852 9.670270603
Table 5. Inhibitor screen shows that G7, G6, A9, and I1 are the top inhibitors. Three trials were
conducted for each substrate and we averaged the three absorbance values. Utilizing a standard curve of
the absorbance at 605 nm versus [Phosphate] created using given spectrophotometric data, absorbances
were converted into final concentration of phosphate.
25
(e)
Figure 8. Changes to the G6 inhibitor significantly increase the affinity of the inhibitor to VlCdc14.
(a). Original inhibitor in binding pocket displays an affinity score of -6.53 kcal/mol. The original
inhibitor is G6 from the inhibitor screen, and it was chosen because of its consistent binding of the active
site and relatively high percent-inhibition. (b). Original Inhibitor Ligand Interaction Map shows
interaction between inhibitor’s sulfur trioxide group and Arg341. Refer to figure 8(e) for the legend.
(c). Optimized inhibitor in binding pocket displays an affinity score of -9.58 kcal/mol. Modifications
include the addition of two amine groups, two hydroxylated fluorine atoms, an imidazole group, and an
hydroxyl group (d). Optimized Inhibitor Ligand Interaction Map Shows interactions with Cys335
26
and Arg341. Cystine-355 and Arginine-341 were shown to be involved in the enzyme mechanism, so an
effective competitive inhibitor should interact with these residues. (e). Legend for the Ligand
Interaction Map
27