Functional characterization of Candida albicans Hos2 histone deacetylase [version 3; peer review: 1 approved, 2 approved with reservations]

Candida albicans is a mucosal commensal organism capable of causing superficial (oral and vaginal thrush) infections in immune normal hosts, but is a major pathogen causing systemic and mucosal infections in immunocompromised individuals. Azoles have been very effective anti-fungal agents and the mainstay in treating opportunistic mold and yeast infections. Azole resistant strains have emerged compromising the utility of this class of drugs. It has been shown that azole resistance can be reversed by the co-administration of a histone deacetylase (HDAC) inhibitor, suggesting that resistance is mediated by epigenetic mechanisms possibly involving Hos2, a fungal deacetylase. We report here the cloning and functional characterization of HOS2 (H igh O smolarity S ensitive) , a gene coding for fungal histone deacetylase from C. albicans . Inhibition studies showed that Hos2 is susceptible to pan inhibitors such as trichostatin A (TSA) and suberoylanilide hydroxamic acid (SAHA), but is not inhibited by class I inhibitors such as MS-275. This in vitro enzymatic assay, which is amenable to high throughput could be used for screening potent fungal Hos2 inhibitors that could be a potential anti-fungal adjuvant. Purified Hos2 protein consistently deacetylated tubulins, rather than histones from TSA-treated cells. Hos2 has been reported to be a putative NAD+ dependent histone deacetylase, a feature of sirtuins. We assayed for sirtuin activation with resveratrol and purified Hos2 protein and did not find any sirtuin activity. with anti-acetylated Histone (Ac-K-9) the deacetylation of by rHos2. assay in three independent deacetylase This manuscript describes an in vitro assay strategy for testing the effects of antifungal compounds on the activity of Hos2 deacetylase large scale screens.The goal is to find compounds that could be used as facilitators in combinatorial drug treatments for azole resistant Candida albicans infections. Although in vitro assays are sometimes useful for developing in vivo strategies, they are not ideal for screening and identifying optimal drugs. As shown in this study by the authors, the recombinant Hos2 they generated exhibits paradoxically tubulin-specific deacetylase activity and not Hos2 specific activity. This finding seems to undermine their intent, because anti-tubulin deacetylase activity or anti-human HDAC6 activity demonstrated in vitro could be an artifact of the recombinant protein. In addition, anti-tubulin and anti-HDAC6 activity might result in adverse side effects on human host functions and thus affect the specificity of the primary drug. As a pharmaceutical organization, involved in drug discovery, our initial compound libraries are screened for in vitro activity/inhibition assays. These in vitro assays act as important filters removing in-active compounds to progress only the most potent ones to in vivo screening. Deacetylase assays using recombinant Hos2 enzyme and synthetic substrate Boc (Lys) AMC showed that the enzyme is active in standard fluorogenic assays. Absence of deacetylase activity with histone preparations (Human/ Candida ) in the in vitro histone deacetylation assays was a surprise but the fact that Class I HDAC inhibitor like MS-275 did not inhibit rHos2 in standard fluorogenic assays, led us to explore alternate substrate, namely tubulins. This is an interesting study designed to examine the biological activities of Candida albicans Hos2 enzyme, a putative histone deacetylase. The authors express a recombinant version of the protein, which has a His tag added to the N-terminus, and use this to explore potential substrates for the enzyme.

This fungus is of clinical importance and is one of the leading causes of systemic infections in immunocompromised individuals. C. albicans is the fourth most common cause of nosocomial bloodstream infections and is associated with high mortality rates 2 . Azoles and echinocandins targeting the ergosterol and cell wall biosynthesis pathway respectively have been used as anti-fungal drugs though the emergence of drug-resistant strains has compromised the efficacy and utility of these drugs 3 .
Azole resistance in Candida sp. is mediated by up regulation of genes encoding ERG11, a lanosterol demethylase 4-6 , MDR1 5,7 and by CDR (Candida drug resistance) efflux pumps 5,7-9 . A combination of existing anti-fungals with new classes of drugs that act by different mechanisms will be viable alternatives to the current monotherapy regimen, which contributes to the emergence of drug resistance.
Histone deacetylases (HDAC) play an important role in modulating chromatin conformation, by deacetylating crucial lysine residues in the histone octamers over which the chromatin DNA are wrapped 10 . Human HDAC`s fall into four broad categories, Class I (HDAC1, 2, 3, and 8), Class II a (HDAC 4, 5, 7 and 9) Class II b (HDAC 6 and 10) Class III (sirtuins) and Class IV (HDAC11) based on sequence homology, substrate preference and co-factor requirements. The involvement of each of these isoforms in disease pathology has been elucidated to some extent in recent times. The approval of suberoylanilide hydroxamic acid (SAHA) 11 , a well known inhibitor of HDACs by the US FDA for treating CTCL, (cutaneous T cell lymphoma) 12,13 has thrown open the doors for exploring the use of HDAC inhibitors in combination with existing drugs for several diseases, such as malaria and Kala-azar etc., [14][15][16] .
Class specific inhibitors are now becoming a reality for human HDAC isoforms. For example the HDAC Class I specific inhibitor MS-275 is in advanced clinical trials (clinical trial Nos. NCT00020579, NCT00866333) for several forms of cancer, and the HDAC Class II specific inhibitor ACY-1215 is at an advanced clinical phase (clinical trial Nos. NCT01323751, NCT01583283) for myeloma.
HDAC inhibitors have been shown to synergize the actions of antifungal agents, due to their effect on preventing drug resistance in vitro 17,18 . Therefore, an alternative approach to address fungal drug resistance could be to harness the potential of modulating fungal gene expression by inhibition of fungal HDACs 17,19 . Cloning and expression of HOS2 in an insect cell expression system Oligos were designed with codon changes made for 4 th and 271 st serine residues. Full length HOS2 gene was amplified by using 4 different primers (Table 1) using splicing by overlap extension (SOE PCR), so that codon usage could be maintained in any heterologous expression system. The full-length blunt end PCR product was cloned in to pJET1.2 cloning vector (CloneJET PCR Cloning Kit Thermo Scientific) and confirmed by restriction digestion using BamH1 and Not1. DNA sequencing using T7 promoter (forward,

Amendments from Version 2
In this version of the manuscript, we have made some minor corrections as suggested by Referee 3. In the result section, the codon sequence for amino acid at 4 th and 271 st position in native HOS2 gene is mentioned. These codons were mutated to express serine using the oligonucleotide primers, where the nucleotides responsible are made to appear bold, red colored and underlined in Table 1.

REVISED
5´-TAATAC GACTCACTATAGGG-3´) and pJET1.2 (reverse, 5´-AAGAACATCGATTTTCCATGGCAG-´3) sequencing primer ascertained the sequence of the recombinant Hos2 gene. Production of polyclonal anti sera against Hos2 protein Female BALB/c mice were purchased from The Jackson Laboratory (Bar Harbor, ME) at 4 weeks of age and then housed at the animal facility, Orchid Chemicals and Pharmaceuticals for 2 weeks in a specific-pathogen free facility with a 12 h light cycle (6 am-6 pm) and a 12 h dark cycle (6 pm-6 am). Groups of four mice were housed in sterilised polypropylene cages covered with stainless steel grid top, lined with autoclaved clean rice husk bedding. All animal experimentations were approved by the institutional animal ethics committee (Protocol No. 01/IAEC-05/PPK/2009). The native Hos2 protein (expressed in pET-32 bacterial vector system and purified using nickel affinity chromatography under denaturating conditions) was emulsified in complete Freund's adjuvant and injected subcutaneously into two female BALB/c mice (20 μg/mice). Booster doses of the deacetylase antigen (20 μg/mice) were given on the 14 th and 21 st day in Freund's incomplete adjuvant. Mice were bled 7 days after the second booster dose and polyclonal anti sera was separated after clotting the blood 26 .
Cell culture Jurkat, a human T lymphocyte cell line, and HeLa, a human cervical adenocarcinoma cell line was obtained from ATCC and were cultured in DMEM (Gibco, Life technologies) supplemented with 10% (v/v) fetal calf serum (FCS), 2 mM glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin (Gibco, Life technologies).
Isolation of nuclear histones from mammalian cells Acetylated histones were isolated from HeLa cells treated with the HDAC inhibitor SAHA as per published protocol 27 . The histone pellet was then resuspended in ultra pure water, stored in 50 μl aliquots at -70°C and the protein concentration was determined using a BCA kit (Pierce).

Isolation of histones from Candida sp.
C. albicans ATCC 90028 mycelia (~5 gm wet weight) were washed with water, centrifuged at 10,000 rpm for 10 minutes at 4°C and the mycelial pellet was resuspended in 50 ml of 0.1 mM Tris-HCl, pH 9.4, 10 mM DTT. The sample was incubated with shaking at 30°C for 15 min and pelleted. The pellet was washed with 50 ml of sorbitol/HEPES buffer (1.2 M sorbitol, 20 mM Hepes, pH 7.4) and left resuspended in the same buffer containing lyticase (1000 units) overnight at 30°C for spheroplasting 28 . The sample was pelleted and proceeded to histone isolation using acid extraction as described previously 29 . Acetone was added at 3:1 (vol/vol) to precipitate the histones, which were subsequently dissolved in 10 mM Tris, pH-8.0.

Deacetylation of nuclear histones
Acetylated histones isolated from SAHA treated HeLa cells were used for the deacetylation assay with recombinant Hos2 enzyme.
In brief, purified acetylated histones (2 μg) were incubated with different amounts of recombinant Hos2 enzyme. Histone deacetylation with 300 ng of rhHDAC1 or rhHDAC6 (expressed in-house using Baculo-viral expression system) was used as a positive control. Deacetylation assays were carried out in 100 μl reaction volume for 1 hr at 37°C in reaction buffer (50 mM Tris Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 2.5 mM MgCl 2 , 1 mg/ml BSA). At the end of incubation, the reaction was stopped by the addition of 1X Lammeli sample buffer. The protein samples were resolved by SDS-PAGE and immunoblotted with anti-acetylated H3 Histone (Ac-K-9) antibody to study the deacetylation of H3-histone by rHos2. The assay results are reproducible in three independent experiments.

Deacetylation of acetylated tubulin
Whole cell extracts from TSA (Sigma, St. Louis, MO, USA) treated Jurkat cells were used for the α-tubulin deacetylation assay with recombinant Hos2 enzyme. In brief, whole cell extract (10 μg) were incubated with 5 and 8 μg of recombinant Hos2 protein. Deacetylation assays were carried out in 100 μl reaction volume for 3 hr at 37°C in reaction buffer (50 mM Tris-Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 2.5 mM MgCl 2 ). At the end of incubation, the reaction was stopped by the addition of 1X Lammeli sample buffer. The protein samples were resolved by SDS-PAGE and immunoblotted with either anti-acetylated α-tubulin or anti-α-tubulin antibodies. The assay results are reproducible in two independent experiments.

Statistical analysis
All values are expressed as mean ± standard deviation and the graphs were generated using Graph-Pad Prism ® (Version 4) for Windows (GraphPad Software, San Diego, California, USA. Statistical analysis was performed by one-way analysis of variance (ANOVA), followed by Bonferroni multiple comparison test for all parameters.

HOS2 gene PCR
Codon usage in C. albicans is different from standard genetic code. The 4 th serine (CUG) and the 271 st serine (CUG) of HOS2 are translated as leucine in both mammalian and in insect cell expression systems. Hence, these codons were mutated using oligonucleotide primers to express serine in the recombinant Hos2 protein. The PCR product was cloned in the pFastbac-HTB shuttle vector and subsequently into baculo viral DNA using the Bac-to-Bac expression system.

Protein expression and purification
The Hos2 enzyme was expressed in the baculoviral-insect cell expression system as a NH 2 -terminal hexa histidine tagged fusion protein, which is detected on Western blot as a ~52 kDa protein using our own polyclonal anti-Hos2 anti-sera raised against Hos2 protein in mice. SDS-PAGE analysis of purified protein revealed a major band at ~52 kDa ( Figure 1A). In vitro deacetylation assay using synthetic peptide substrate Recombinant Hos2 enzyme was assayed for deacetylase activity using the synthetic deacetylase substrate, Boc-Lys (ac)-AMC. The total activity with Boc-Lys (ac) AMC showed the enzyme to be active in deacetylating the lysine residue and the activity increased significantly (P < 0.05) with an increase in Hos2 concentration ( Figure 1B).
The inhibition of deacetylation activity of recombinant Hos2 was studied using classical HDAC inhibitors namely SAHA, TSA and MS-275. TSA was very potent in inhibiting Hos2 with an IC 50 of 2.8 nM, SAHA inhibited Hos2 with an IC 50 of 65 nM (Figure 2, Table 2). However, MS-275 showed > 50% inhibition of Hos2 activity only at 10 μM (Table 3).
In vitro deacetylation assay using natural substrates The ability of purified Hos2 protein to deacetylate acetylated histones was examined in vitro using acetylated nuclear histone preparation made from SAHA treated HeLa cells. The nuclear histones from HeLa cells were isolated using a modified protocol of Shechter et al. 27 and established the deacetylation assay using rhHDAC1/ rhHDAC6 as controls along with rHos2. In these assays it was found that 0.3 μg of rhHDAC1 was able to deacetylate the nuclear acetylated histones as detected by an anti-H3-K9 histone antibody. Recombinant hHDAC1 was more potent in deacetylating lysine   residues in H3-histones than rhHDAC6. However, no significant deacetylation of H3-Histone was seen with 0.3 μg of Hos2 ( Figure 3A). Similar results were obtained with acetylated histones from Candida in the in vitro histone deacetylation assay with purified Hos2 protein (up to 3 μg, Figure 3B). In contrast, when recombinant Hos2 was incubated with mammalian acetylated α-tubulin, a significant reduction in acetylation of α-tubulin (K40) could be observed (Figure 4). In vitro Sirt1 deacetylation assay using fluor-de-lys substrate Since Hos2 is a putative NAD+ dependent deacetylase, a feature of sirtuin class of deacetylases, it was of interest to check if the Hos2 protein displayed any sirtuin like activity. The sirtuin-like activity in response to the sirtuin activator resveratrol was studied using fluor-de-lys, a synthetic Sirt1 substrate. HeLa nuclear extract was used as a positive control for sirtuin activity. In the presence of TSA where the HDAC activities are inhibited no appreciable NAD+ dependent deacetylase-like activity was seen, following incubation of Hos2 with different concentrations of resveratrol ( Figure 5).

Discussion
Pathogenic fungi are increasingly responsible for life threatening infections in the elderly and immunocompromised patients. While some species have intrinsic resistance to anti-fungals, others develop resistance during the course of treatment. Increasing antifungal resistance and treatment failures in patients is becoming a challenge.
The Candida genome encodes at least 3 distinct classes of histone deacetylases in addition to sirtuins. There are 8 different histone deacetylases (HOS1, HOS2, HOS3, HDA1, HDA2, HDA3, RPD3, It has been surmised that inhibiting Hda1 for example might enhance the anti-fungal effect of HDAC inhibitors by limiting hyphal development, while inhibiting Hos2 might contribute to limiting yeast development 32 . Hos2 has an essential function in morphogenesis especially during conditions of nitrogen starvation 33 . The critical role of HDAC`s in C. albicans pathogenesis and survival to antifungal treatment underscores the necessity to study HDAC function in this organism. The increasing clinical incidences of azole resistant fungal infections in critical care patients, makes a good reason to find additional drug targets to control such diseases. There is at least one small molecule (MGCD 290) that inhibits Hos2 histone deacetylase that has progressed to clinical trials. MGCD 290 in combination with azoles was shown to be active against azole resistant yeasts and moulds 18 .
In order to better understand the role played by the Candida Hos2 enzyme we attempted to clone, express and characterize the protein in detail. This study describes the cloning, expression, purification and characterization of the Hos2 deacetylase enzyme from C. albicans.
We cloned and expressed the HOS2 gene in baculoviral expression system as a 6x his-tagged protein, which exhibits classical deacetylase activity with the synthetic Boc-Lys (ac)-AMC peptide substrate. In our study, the yield of the Hos2 protein was generally low and probably could be attributed to difference in codon usage between Candida and Sf9 insect cells. This in vitro enzymatic assay, amenable to high throughput, could be used for screening potent fungal Hos2 inhibitors that could be a potential anti-fungal adjuvant.
Our studies with the recombinant Hos2 protein showed that it is susceptible to inhibition by standard HDAC inhibitors such as SAHA and TSA. We characterized the inhibition profile of purified proteins with SAHA and TSA and showed that TSA is a more potent inhibitor of Hos2 with an IC 50 of 2.8 ± 0.9 nM compared to SAHA (IC 50 65.4 ± 2.4 nM). Our studies with the Class I HDAC inhibitor MS-275 showed that this inhibitor did not inhibit Hos2 deacetylase as effectively as the pan HDAC inhibitors SAHA or TSA, suggesting that Candida Hos2 is more similar to Class II deacetylases.
The recombinant Hos2 failed to deacetylate either mammalian or fungal nuclear histones, suggesting that the histones are not the preferred substrates for the Hos2 enzyme. The fact that, the recombinant Hos2 enzyme did not show any inhibition with the Class I inhibitor MS-275 led us to explore alternate substrates including tubulins, which are substrates for Class II histone deacetylases. Experiments with total lysates from Jurkat cells containing acetylated α-tubulin showed a dose dependent deacetylation albeit at higher concentration of Hos2 (> 5 μg). Hos2 in essence resembles the Class II mammalian HDACs, specifically HDAC6 in its preference for tubulin deacetylation. It has been shown that microtubules in the fungal hyphae drive nuclear dynamics and cell cycle progression to morphogenesis 34 . In view of the fact that Hos2 seems to preferentially deacetylate tubulins, it would be interesting to see if Hos2 inhibitors would act as anti-fungals, either as a monotherapy or in synergy, with existing anti-tubulin agents such as benomyl, nocodazole etc. The physiological relevance of tubulin deacetylation by Hos2 warrants further study.
The Candida genome database 23 predicts Hos2 protein to be a NAD+ dependent deacetylase. Sirtuins which are classified as Class III deacetylases are NAD+ dependent enzymes activated by polyphenols such as resveratrol. Sirtuins have been proposed to regulate cellular metabolism, ageing and other related processes, specifically cellular stress response to caloric restriction, mediating life span extension. The role of resveratrol as a sirtuin activator has been resolved recently and it is now known that in addition to activating Sirt1, it also activates Sirt5, while inhibiting Sirt3 25 . Thus inhibiting any sirtuin like activity with small molecule inhibitors could be another way of enhancing the activity of currently used anti-fungals. We evaluated the possibility of Hos2 being a sirtuin like enzyme with a known Sirt1 activator resveratrol. We did not observe any significant (P value 0.5317 and 0.4411, in the presence and absence of trichostatin respectively) activation of NAD+ dependent deacetylase activity with the fluor-de-lys substrate.
In conclusion this study establishes a functional assay for purified Hos2 protein. This in vitro enzymatic assay can be used to screen small molecule inhibitors of Hos2, which can synergise current anti-fungals in the clinic.

Competing interests
No competing interests were disclosed.

Grant information
This was an exploratory in-house project funded solely by Orchid Chemicals and Pharmaceuticals Limited.
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.  The points raised by the referee are addressed in the new version of the article.

Open Peer Review
We have carried out a separate experiment to show that the signal in the western blot is specific (this can be downloaded here but is not included in the revised article).
Fig 1B was specifically modified at the request of the editorial team during the initial stage and subsequently modified based on inputs from other referee(s).

David Soll
Department of Biology, University of Iowa, Iowa City, IA, USA This manuscript describes an in vitro assay strategy for testing the effects of antifungal compounds on the activity of Hos2 deacetylase large scale screens.The goal is to find compounds that could be used as facilitators in combinatorial drug treatments for azole resistant Candida albicans infections. Although in vitro assays are sometimes useful for developing in vivo strategies, they are not ideal for screening and identifying optimal drugs.
As shown in this study by the authors, the recombinant Hos2 they generated exhibits paradoxically tubulin-specific deacetylase activity and not Hos2 specific activity. This finding seems to undermine their intent, because anti-tubulin deacetylase activity or anti-human HDAC6 activity demonstrated in vitro could be an artifact of the recombinant protein. In addition, anti-tubulin and anti-HDAC6 activity might result in adverse side effects on human host functions and thus affect the specificity of the primary drug.
It is surprising that the recombinant Hos2 did not show any activity in acetylating either the human or Candida histone H3 in vitro, despite the fact that several earlier studies in C. albicans has demonstrated that Hos2 is involved in morphogenetic programs by affecting chromatin function.
In addition, an isolated study has demonstrated the efficacy of a Hos2-specfic inhibitor on azoleresistant strains of C. albicans (Pfaller et al., 2009), which is quoted by the authors. The authors do state in the introduction that Hos2 is not a typical deacetylase and has been described by the Candida Genome Database as a member of NAD-dependent sirtuins. The authors' data show that their recombinant version lacks this activity, but does exhibit inhibition by classical inhibitors of histone deacetylases.
In addition to this major problem there are a few other aspects of the paper that need to be addressed: Firstly, Figure 1 and Figure 5 do not have error bars to indicate reproducibility of the data. Also, a statistical analysis of data sets would be useful to indicate the significance of the results.
○ Secondly, the units in Figure 2 and Table 2 need to be consistent with the units described in the text.

○
Thirdly, since Figure 4 shows the effect of recombinant Hos2 on only tubulin, the title needs to be changed to indicate that the assay was not done with purified histones. ○ Finally, it is not clear what "fold activity" (in Figure 1B) and "fold activation" (Figure 5) refer to. I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

© 2014 MacCallum D.
This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Donna MacCallum
Aberdeen Fungal Group, University of Aberdeen, Aberdeen, UK This is an interesting study designed to examine the biological activities of Candida albicans Hos2 enzyme, a putative histone deacetylase. The authors express a recombinant version of the protein, which has a His tag added to the N-terminus, and use this to explore potential substrates for the enzyme.
There are a number of major issues with the data as presented, which need to be addressed: It is difficult to make any conclusions from the quantitative data presented in the figures as there are no error bars and the number of replicates is not stated. In addition, it is unclear what the data in Figures 1B and 5 has been expressed relative to (i.e. fold change relative to...?) Statistical analyses should also be carried out to determine whether any differences are statistically significant ones.

○
The authors should also revisit the data presented in Figure 2 and Table 2. The data does not correlate, as the authors suggest that the IC 50 values are in micromolar amounts, yet the data in the figure would suggest millimolar levels -units should be checked.

○
The title for Figure 4 should be modified, since this figure contains only data for betatubulin.

○
In the main text the authors need to be much clearer regarding the rationale for looking at tubulin as a substrate. The authors may also wish to consider the possibility that the substrate specificity and/or activity may have been affected by the N-terminal tag and how this could be investigated.

○
The discussion section should be revisited to remove the reiteration of the introduction and results. Instead, use this part of the paper to discuss the findings in relation to other published work.

Minor points:
Candida albicans is capable of causing superficial (oral and vaginal thrush) infections in immune normal hosts (abstract and introduction). ○ Gene names, e.g. CDR1 and ERG11, should be in italics.

○
For the enzymatic assay it would be good to state the volume of the assay so that final concentrations can be worked out. I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.