Characterization of the Pneumocystis jirovecii and Pneumocystis murina phosphoglucomutases (Pgm2s): a potential target for Pneumocystis therapy

ABSTRACT

Pneumocystis cyst life forms contain abundant β-glucan carbohydrates, synthesized using β-1,3 and β-1,6 glucan synthase enzymes and the donor uridine diphosphate (UDP)-glucose. In yeast, phosphoglucomutase (PGM) plays a crucial role in carbohydrate metabolism by interconverting glucose 1-phosphate and glucose 6-phosphate, a vital step in UDP pools for β-glucan cell wall formation. This pathway has not yet been defined in Pneumocystis. Herein, we surveyed the Pneumocystis jirovecii and Pneumocystis murina genomes, which predicted a homolog of the Saccharomyces cerevisiae major PGM enzyme. Furthermore, we show that PjPgm2p and PmPgm2p function similarly to the yeast counterpart. When both Pneumocystis pgm2 homologs are heterologously expressed in S. cerevisiae pgm2Δ cells, both genes can restore growth and sedimentation rates to wild-type levels. Additionally, we demonstrate that yeast pgm2Δ cell lysates expressing the two Pneumocystis pgm2 transcripts individually can restore PGM activities significantly altered in the yeast pgm2Δ strain. The addition of lithium, a competitive inhibitor of yeast PGM activity, significantly reduces PGM activity. Next, we tested the effects of lithium on P. murina viability ex vivo and found the compound displays significant anti-Pneumocystis activity. Finally, we demonstrate that a para-aryl derivative (ISFP10) with known inhibitory activity against the Aspergillus fumigatus PGM protein and exhibiting 50-fold selectivity over the human PGM enzyme homolog can also significantly reduce Pmpgm2 activity in vitro. Collectively, our data genetically and functionally validate phosphoglucomutases in both P. jirovecii and P. murina and suggest the potential of this protein as a selective therapeutic target for individuals with Pneumocystis pneumonia.

INTRODUCTION

Every year, more than 400,000 individuals worldwide suffer infections from Pneumocystis jirovecii, the causative agent for Pneumocystis jirovecii pneumonia (PJP) (1). In HIV and other immunosuppressive states, the absence of effective lymphocytic immunity results in an exuberant and often fatal Pneumocystis infection (2). Even though the advent of highly active antiretroviral therapy (HAART) has significantly reduced morbidity and mortality in individuals with AIDS, PJP remains a significant health threat to immunocompromised individuals with AIDS (3, 4). Overall, recent estimates of the total healthcare cost due to PJP are $475–$686 million annually (5).

The Pneumocystis cyst cell wall contains abundant amounts of β-1,3 and β-1,6 glucans, providing rigidity and mechanical strength for the organism to survive outside the host (6–8). Other researchers have demonstrated the essential role of the cyst life form, which produces eight daughter nuclei (pre-trophic forms), in the progression of the life cycle and infectivity within the host’s lungs (9, 10). The absence of these carbohydrates in humans has made the fungal cell wall an attractive therapeutic target. The synthesis of fungal β-glucans relies on a continuous supply of UDP-glucose (UDP-Glc) donor pools (11). Phosphoglucomutase (PGM) is a key enzyme involved in glucose metabolism, catalyzing the interconversion of glucose 1-phosphate (Glc-1-P) and glucose 6-phosphate (Glc-6-P). Additionally, UDP-Glc pyrophosphoglucomutase converts Glc-1-P and UTP to UDP-Glc and pyrophosphate. The activity of PGM can be inhibited by lithium through the competitive displacement of Mg₂⁺ from its binding site (12, 13), resulting in reduced enzymatic activity (13, 14). An illustration showing the major metabolic pathways with PGM involvement is shown in adapted Fig. 1 (15). Furthermore, others have shown recently, via a thio-reactive compound screen, an isothiazolone fragment targeting a cysteine residue not conserved in the human ortholog. Furthermore, via the implementation of scaffold synthesis, these authors created a para-derivative (ISFP10) of this compound with an IC₅₀ of 2 μM against Aspergillus fumigatus PGM and 50-fold selectivity over the human PGM ortholog (16). Therefore, selective inhibition of PGM activity might be a viable therapeutic antifungal target effecting Pneumocystis growth and/or viability.

Fig 1.

In eukaryotic cells, phosphoglucomutase (PGM) is vital to major metabolic pathways. Glc, glucose; Gal-1-P, galactose-1-phosphate; Glc-1-P, glucose-1-phosphate; Glc-6-P, glucose-6-phosphate. Figure adapted from Csutora et al. (15).

Here, we demonstrate that both Pneumocystis jirovecii and Pneumocystis murina possess functional PGM enzymes. By utilizing heterologous expression, we show that Pjpgm2 and Pmpgm2 restore growth in liquid and solid media containing galactose. Furthermore, pgm2Δ Saccharomyces cerevisiae cells exhibit significantly higher sedimentation rates compared to the wild-type parent strain, and the expression of Pjpgm2 and Pmpgm2 individually in this strain can restore sedimentation rates to wild-type levels. In vitro, we demonstrate that the expression of both Pjpgm2 and Pmpgm2 facilitates the restoration of PGM activity in pgm2-null yeast. This activity is significantly repressed in the presence of lithium. Next, we demonstrate that the use of lithium can significantly reduce P. murina cell viability in dose fashion, similar to the effects of the well-established anti-Pneumocystis drug pentamidine. Finally, we show that ISFP10, a previously published inhibitor against A. fumigatus PGM with approximately 50-fold selectivity over the human PGM enzyme, can also significantly reduce Pneumocystis PGM activity in vitro. These results collectively establish that both PjPgm2p and PmPgm2p are functional PGMs, making this protein a potential attractive therapeutic target for individuals with PJP.

MATERIALS AND METHODS

Media and strains

The S. cerevisiae wild-type (BY4742) and pgm2Δ (YMR105C) strains were obtained from Horizon Discovery (Waterbeach, UK). All S. cerevisiae cultures were grown at 30°C in yeast extract-peptone-dextrose medium or minimal medium containing 2% glucose or 2% galactose and lacking uracil. Escherichia coli One Shot TOP10F' chemically competent cells (Invitrogen, Waltham, MA) were used for TOPO TA cloning and bacterial transformation following the manufacturer’s protocol. All E. coli cultures were grown overnight at 37°C.

For the yeast complementation assays described later, we utilized the pYES2.1 TOPO TA Expression Kit. cDNAs are expressed (induced) by the removal of glucose-containing media and the addition of media with galactose. The expression control plasmid pYES2.1/V5-His/lacZ was included as a control.

Isolation of P. murina

For RNA isolation and downstream cDNA generation, P. murina organisms were derived from the American Type Culture Collection, strain PRA-111 (ATCC, Manassas, VA), and propagated through Rag2^tm1FwaIl2rg^tm1Wjl mice (purchased from Taconic Biosciences) (17, 18). Animals were infected with P. murina intratracheally, and after 8 weeks of infection, organisms were harvested as previously noted (19, 20).

Cloning and sequence analysis of PjPgm2 and Pmpgm2

To determine if P. jirovecii and P. murina contain potential phosphoglucomutases (PGMs), we performed a keyword search of the NIH BLASTP Basic Local Alignment Search Tool (BLAST) with the S. cerevisiae Pgm2 protein homolog using the P. jirovecii and P. murina non-redundant translated protein sequences. Complete mRNA sequences of PjPgm2 (accession #XM_018373800.1) (T551_01537) and Pmpgm2 (accession #XM_007874716) (PNEG_00993) were obtained from the National Institutes of Health (NIH) Pneumocystis jirovecii RU7 and Pneumocystis murina B123 genome sequencing projects (21). Primers for Pjpgm2 and Pmpgm2 were synthesized to generate full-length cDNA sequences (Table 1). The predicted amino acid sequences of PjPgm2p and PmPgm2p were compared to the S. cerevisiae Pgm2p enzyme, respectively, using MacVector version 18.6.1 (22) to generate protein sequence alignments.

TABLE 1.

Primer sets used for cDNA generation and qPCR

Gene name	Forward primer	Reverse primer
Pmpgm2	ATGAATATAGAAACAGTACAAAGCC	TTATGTTATTACTGTAGGAGTGCAAG
Pjpgm2	ATGCATGTTCAAGCAGTACAAACC	TTATGTTATGACCGTAGGTGTGC
Pm16S mitochondrial ribosomal RNA	GATGGCTGTTTCCAAGCCCA	GTGTACGTTGAAAGTACTC

Functional complementation of S. cerevisiae strains

The complete full-length Pjpgm2 and Pmpgm2 cDNAs were amplified utilizing a synthetically generated full-length PjPgm2 cDNA created by Integrated DNA Technologies as a template or from P. murina cDNA (generated by reverse transcription of RNA using the Bio-Rad iScript Advanced cDNA Synthesis Kit) using Platinum Taq DNA Polymerase High Fidelity (Invitrogen) using the PCR primers listed in Table 1. After ligation to the pYES2.1 TOPO TA yeast expression vector (Invitrogen), the plasmid was sequenced and transformed into E. coli One Shot TOP10F' chemically competent cells. Next, this expression plasmid was cloned into a pgm2-deficient yeast strain (pgm2Δ) (YMR105C). This knockout strain, along with parent strain BY4742, was also transformed with the control vector pYES2.1/V5-His/lacZ according to the manufacturer’s protocol. For complementation studies, after overnight growth of the cells in liquid minimal medium lacking uracil and containing 2% galactose or glucose, the transformants were serially diluted 1:10 (starting at an OD600nm of 0.3) and plated on synthetic medium with agar minus uracil, with and containing 2% galactose or glucose, and, where shown, 30 mM lithium chloride (LiCl). Plates were incubated at 30°C for 2 days and photographed.

Yeast growth curves

The yeast strains were cultured in 10 mL of minimal media without uracil and 2% glucose for 18 h at 30°C. The next day, the OD600_nm of the yeast cells was measured, and appropriate volumes were taken to achieve a starting OD600_nm of 0.4 in 15 mL of minimal media without uracil and with 2% galactose. The yeast cells were then pelleted at 2,500 RPM for 5 min, and the cultures were grown at 30°C and 225 RPM. The OD600_nm was determined at 18 h.

Sedimentation assays

The yeast strains in minimal media (10 mL) without uracil and 2% galactose were grown for 18 h at 30°C. The next day, the yeast strains were diluted to an OD600_nm of 1.0 in 1-mL cuvettes. Every hour for an 8-h total, the undisturbed cuvette’s 0D600_nm was read.

Analysis of PjPgm2p and PmPgm2p enzyme activity

The plasmid pYES2.1/V5-His-TOPO (Invitrogen) was used for cloning the Pjpgm2 or Pmpgm2 cDNA construct into E. coli One Shot TOP10F' chemically competent cells for sequencing. Once the correct in-frame cDNA sequences were confirmed, the plasmid was transformed into the pgm2Δ S. cerevisiae strain to express the respective PGM proteins. This plasmid contains the galactose 1 (GAL1) promoter to induce protein expression in the presence of galactose. Yeast cells were grown overnight in a minimal liquid medium lacking uracil and containing 2% glucose at 30°C. The following day, the OD600_nm was determined, and the cultures were diluted to an OD600_nm of 0.4 in 15 mL of minimal liquid medium lacking uracil and containing 2% galactose at 30°C and grown for 18 h. Next, the cells were collected by centrifugation at 3,000 × g for 5 min. Crude extracts from S. cerevisiae cells were obtained by a combination of treatment with Y-PER lysis reagent (Thermo Fisher Scientific, Waltham, MA) and TissueLyser LT with 7-mm stainless steel beads for 5 min at 50 oscillations per second (Qiagen, Germantown, MD) (18) in the presence of complete protease inhibitor cocktail tablets (Sigma-Aldrich, St. Louis, MO). Insoluble material and cell wall debris were removed by centrifugation, and the total protein was determined using the Pierce BCA protein assay kit. Aliquots were stored at −20°C until use. The enzymatic activity of PjPgm2p and PmPgm2p in transformed yeast extracts was assessed similarly to that previously described (23) using 2.0 µg yeast protein assayed by the PGM activity system from Sigma-Aldrich. To determine the potential effects of lithium on PGM activity, 200 µM (14) was used in the assays noted. Lastly, the same assay conditions noted above were used with the A. fumigatus PGM-specific inhibitor para-aryl derivative ISPF10 compound at 100 and 150 µM, respectively, and maintained in 3% DMSO (16). For all assay readouts, production formation was measured by the formation of nicotinamide adenine dinucleotide, reduced form, as measured by the change in absorbance at 450_nm at 5 min.

Effects of lithium on Pneumocystis viability

Previous reports note that lithium also has anti-fungal activities (14, 24–26) and acts as a competitive inhibitor of yeast phosphoglucomutase activity (14). Therefore, we tested LiCl (0.5 and 1 mM) (25) with freshly isolated P. murina organisms maintained ex vivo in viability medium for 24 h. Pentamidine isethionate (Sigma-Aldrich) was used at 5 and 10 µM as a positive control (27). Relative viability was assessed by measuring P. murina 16S mitochondrial ribosomal copy number as previously noted (28). Approximately 5 × 10⁷ total organisms (10:1 ratio of trophic forms to cyst forms) (18) were tested under each condition in 500 µL/well with RPMI 1640 medium containing 20% fetal bovine serum to promote viability. Each compound was tested in duplicate, and a total of four independent experiments were performed. The organisms were incubated at 37°C with 5% CO₂ in standard 24-well plates. After 24 h, equal volumes (50 µL) were removed from each well, total RNA was isolated, and cDNA was made as noted above. qPCR was performed as previously reported (28) utilizing the PCR primers listed in Table 1.

Statistical analyses

The data presented are expressed as mean ± standard deviation. Differences between groups were determined using one-way analysis of variance (ANOVA) and multiple-comparison tests. Graphing and statistical analysis were performed using GraphPad Prism for macOS [version 10.1.0 (264)], with statistical differences considered significant at P values of <0.05, <0.01, <0.001, and <0.0001.

RESULTS

The human and mouse Pneumocystis genomes contain PGM homologs

S. cerevisiae contains two PGM genes, PGM1 and PGM2, with the latter accounting for upwards of 90% of the total cellular PGM activity (29). After conducting BLASTP analysis with the yeast Pgm2 protein, we identified homologs in P. jirovecii and P. murina with 51% and 50% respective identities to the S. cerevisiae strain S288C (NP_013823.1). A comparison of the BLASTP analyses for both P. jirovecii (termed PjPgm2p) and P. murina Pgm2 (termed PmPgm2p) translated proteins showed a 77% primary sequence conservation between both species. These protein alignments are shown in Fig. 2. Interestingly, unlike in S. cerevisiae that contains two PGM homologs, in our protein database search, we only uncovered one PGM gene homolog for both Pneumocystis species.

Fig 2.

Alignment of the predicted Pneumocystis PjPgm2p (strain RU7) and PmPgm2p (strain B123) amino acids with the S. cerevisiae Pgm2 (strain BY4742) protein sequence. Sequence alignments of the Pgm proteins were performed with ClustalW [MacVector, Inc. 18.6.1 (22)] demonstrating significant amino acid homology. Identical residues are shaded in dark gray, and conserved residues are shaded in light gray.

Pjpgm2 and Pmpgm2 restore the growth of pgm2-null yeast in galactose, which is inhibited by lithium

To determine if the Pneumocystis Pjpgm2 and Pmpgm2 genes encode functional PGMs, we tested their ability to complement a S. cerevisiae pgm2-null mutant phenotype. The complete cDNA sequences of Pjpgm2 and Pmpgm2 were expressed individually in the pgm2Δ mutant using the pYES2.1 expression vector. The yeast pgm2Δ strain significantly reduced (~90%) but did not completely void PGM activity due to the presence of a redundant PGM transcript (PGM1) (29). Nevertheless, the mutant strain was severely altered in its ability to metabolize galactose when galactose was used as the only carbon source (30). All four strains tested, including the parent wild type and pgm2Δ with vector control, along with the pgm2Δ constructs containing either the Pjpgm2 (Fig. 3A) or Pmpgm2 (Fig. 3B) cDNAs, exhibited normal growth when glucose was used as the carbon source (Fig. 3A and B, left panel). As expected, when the four yeast strains were grown on galactose, the wild-type strain grew normally, similar to its growth on glucose. However, the pgm2Δ yeast strain with vector control showed impaired growth on galactose. In contrast, the complementation of this strain with either Pjpgm2 or Pmpgm2 completely restored growth (Fig. 3A and B, middle panel). Furthermore, growing the pgm2Δ strain complemented with either Pjpgm2 or Pmpgm2 in the presence of 30 mM LiCl led to significant growth repression, indicating a direct effect of the compound on both PjPgm2p and PmPgm2p PGM activity (Fig. 3A and B, right panel).

Fig 3.

Functional complementation of S. cerevisiae pgm2Δ by the Pneumocystis Pjpgm2 (A) or Pmpgm2 (B) phosphoglucomutase cDNA. Wild-type S. cerevisiae or S. cerevisiae pgm2Δ was transformed with pYES2.1/lacZ control expression vector alone or pYES2.1 + Pjpgm2 or Pmpgm2 cDNAs as indicated. The cells were grown in synthetic medium URA(-) supplemented with either 2% glucose or galactose (wt/vol) as the only carbon source. The last panel shows yeast strains grown in 2% galactose with 30 mM LiCl. The numbers below the OD600_nm are the spectrophotometer readings starting at 0.3 OD600_nm and continuing down as serial 10-fold dilutions of the yeast.

Subsequently, we cultured all four strains in liquid media with galactose as the carbon source for 18 h and measured the final OD600_nm as a measure of growth. As depicted in Fig. 4A and B, complementing the pgm2Δ yeast strain with either PjPgm2 or Pmpgm2 cDNAs resulted in the complete restoration of growth levels to that of the wild-type strain, indicating that the respective Pneumocystis PGMs restored the ability to metabolize galactose in this null mutant (Fig. 4A and B).

Fig 4.

Relative growth curves of wild-type S. cerevisiae or S. cerevisiae pgm2Δ transformed with pYES2.1 control expression vector alone or pYES2.1 + Pjpgm2 (A) or Pmpgm2 (B) cDNAs as indicated. Cell growth was initiated in synthetic medium URA(-) supplemented for 18 h with 2% glucose at 30°C. After 18 h, yeast cells were pelleted and resuspended in synthetic medium URA(-) supplemented with 2% galactose and diluted to an OD600_nm of 0.4. Cells were grown for 18 h with 2% galactose at 30°C, and OD600_nm was determined. Results shown are as means + SEM of three independent experiments; **P < 0.01, ***P < 0.001, ****P < 0.0001.

Previous studies have demonstrated that inhibition of S. cerevisiae pgm2 transcription via the Rim15 gene leads to reduced synthesis of β-glucan and decreased carbohydrate levels in the S. cerevisiae cell wall (31). Furthermore, disruption of β-glucan cell wall synthesis results in reduced alkali-insoluble β-glucans as measured by dry weight (32, 33). Therefore, we measured the sedimentation rate of the four yeast strains after growth in galactose for 18 h, followed by hourly OD600_nm measurements starting at time zero and OD 600_nm of ~1.0 with hourly readings for 8 h. The yeast pgm2-null mutant with vector control displayed significantly greater sedimentation rate over the 8-h time period than the parent wild-type strain. However, the S. cerevisiae pgm2Δ strain complemented individually with either the Pjpgm2 or Pmpgm2 genes restored the sedimentation rate to wild-type parent strain levels, suggesting the importance of the PGMs in Pneumocystis cell wall-glucan synthesis (Fig. 5A and B).

Fig 5.

Relative sedimentation rates displayed by wild-type S. cerevisiae or S. cerevisiae pgm2Δ transformed with pYES2.1/lacZ control expression vector alone or pYES2.1 + Pjpgm2 (A) or Pmpgm2 (B) cDNAs as measured hourly for 8 h. Error bars indicate the standard error of the mean (SEM) for three to five independent experiments. One-way analysis of variance (ANOVA); *P < 0.05.

P. jirovecii and P. murina PGM activity in S. cerevisiae

The activity of P. jirovecii and P. murina respective PGMs was determined in the homogenates of S. cerevisiae cells expressing the respective Pneumocystis enzymes after overnight growth for 18 h. As depicted in Fig. 6A and B, the lysate of pgm2Δ plus vector control cells showed a significant alteration in PGM activity, specifically in converting glucose-1-phosphate to glucose-6-phosphate, compared to the wild-type parent strain plus vector control. The pgm2Δ yeast strain plus vector control still exhibited some PGM activity (~8.7 times less than the parent strain), presumably due to the presence of the minor PGM1 gene (29). However, the pgm2Δ mutant bearing either the PjPgm2 or Pmpgm2 genes individually displayed significantly more PGM activity (~3 and ~6.5×, respectively) than the S. cerevisiae pgm2-null mutant (Fig. 6A and B), demonstrating that the Pjpgm2 and Pmpgm2 genes of P. jirovecii and P. murina encode functional PGM enzymes with the ability to compensate for the severe PGM defect of the pgm2Δ strain.

Fig 6.

Induction of Pneumocystis phosphoglucomutase (PGM) activity in S. cerevisiae pgm2Δ transformed with pYES2.1/lacZ control expression vector alone or pYES2.1 + Pjpgm2 (A) or Pmpgm2 (B) cDNAs. Yeast cells were inoculated in liquid medium and grown overnight at 30°C. The cultures were spun down and then resuspended in synthetic medium containing 2% (wt/vol) galactose at an OD600_nm of 0.4. After overnight growth at 30°C, the yeast cells were collected, and the PGM activity (nM glucose-6-phosphate) in protein homogenates was determined, as described in Materials and Methods. Results shown are as means + SEM of three independent experiments; *P < 0.05, **P < 0.01, ****P < 0.0001, ns, not significant.

Lithium inhibits P. jirovecii PjPgm2p and P. murina PmPgm2p PGM activity

To further determine whether lithium may inhibit PjPgm2p or PmPgm2p enzymes, we measured in vitro PGM activity in the presence of 200 µM LiCl. This LiCl concentration has been shown by others to be the effective IC₅₀ for S. cerevisiae PGM activity (14). Adding LiCl demonstrated significant inhibition of PGM activity of both respective proteins (Fig. 7A and B), indicating that lithium suppresses in vivo PjPgm2p and PmPgm2p PGM activity.

Fig 7.

Lithium inhibits Pneumocystis PGM activity. S. cerevisiae pgm2Δ + pYES2.1 + Pjpgm2 (A) or Pmpgm2 (B) yeast strains were induced as noted previously in synthetic medium containing 2% (wt/vol) galactose at an OD600_nm of 0.4. A cell-free extract of yeast-soluble proteins was assayed in the presence of lithium (200 µM) for PGM activity at a 5-min incubation timepoint. Results shown are as means + SEM of four independent experiments; *P < 0.05.

Effects of lithium on P. murina viability

It has been reported that lithium has anti-fungal activities (14, 24–26) by targeting PGM in S. cerevisiae and Candida albicans (14, 25). To examine whether lithium might alter Pneumocystis viability in culture, we utilized freshly isolated P. murina organisms from mice depleted of CD-4 positive T-cells with the monoclonal GK1.5 antibody. This immunosuppressive method mimics the HIV PJP model (18). We determined that incubating P. murina with either 0.5 or 1 mM LiCl overnight in culture resulted in a significant reduction in fungal viability, as measured by qPCR quantification of the 16S mitochondrial ribosomal copy number, as shown in Fig. 8. This reduction was similar to that observed with the known anti-Pneumocystis agent, pentamidine (Fig. 8) (27), suggesting that lithium may have anti-Pneumocystis properties.

Fig 8.

Lithium reduces P. murina organism viability in vitro. After overnight incubation with lithium or the anti-Pneumocystis drug pentamidine, total fungal RNA was isolated and cDNA was generated for the quantification of 16S mitochondrial ribosomal copy number, as noted in Materials and Methods. Results are expressed from duplicate wells and four independent experiments; *P < 0.05, **P < 0.01.

Isothiazolone ISFP10 inhibits the activity of Pneumocystis PGM

Recently, Yan et al. (16) reported on the first cloning and characterization of the A. fumigatus PGM (AfPGM) enzyme. These researchers demonstrated that AfPGM-directed activity in the fungal organism is required for growth and cell wall integrity. Furthermore, utilizing the Maybridge fragment compound library (1,000 compounds), they identified an isothiazolone fragment as directly targeting a cysteine residue in AfPGM absent from the same location in the human PGM enzyme. Next, through a scaffold synthesis approach, they created a para-aryl derivative (termed ISFP10) with specific activity towards the AfPGM and a reported IC₅₀ value of 2 µM and greater than 50-fold selectivity over the human ortholog (16). Based on these exciting results, we also tested this inhibitor against the PmPgm2 enzyme to see if it could also display activity against the Pneumocystis protein and provide initial in vitro data that, if promising, could be further tested in the in vivo mouse Pneumocystis pneumonia (PCP) model. As described above, we use the glucose-6-phosphate incorporation activity assay to measure the possible effects of the ISPF10 inhibitor on the Pneumocystis PGM at concentrations recommended by the authors, first describing this newly synthesized molecule. As Fig. 9 shows, ISFP10 showed significant inhibitory activity against the Pneumocystis PGM in a dose-responsive manner.

Fig 9.

Isothiazolone ISPF10 inhibits the activity of PmPgm2p. S. cerevisiae pgm2Δ + pYES2.1 + Pmpgm2 cDNA yeast strain was induced as noted previously in synthetic medium containing 2% (wt/vol) galactose at an OD600_nm of 0.4 for 18 h. Cell-free extract of yeast-soluble proteins was assayed in the presence of ISPF10 (100 or 150 µM) for PGM activity at a 5-min incubation period. Results shown are as means + SEM of four independent experiments; *P < 0.05, ***P < 0.001.

DISCUSSION

In 2022, P. jirovecii was included in the recently released World Health Organization (WHO) fungal priority pathogens list (34). Overall, the number of PJP cases has decreased in developed countries over the last few decades due to highly effective antiretroviral therapy. However, there has been a significant increase in cases among non-HIV individuals who undergo immunosuppressive therapies (35, 36). Despite the availability of relatively effective current anti-Pneumocystis therapies, the mortality rate remains high, ranging from 30% to 60% in non-HIV patients and 10% to 20% during initial episodes of PJP in individuals with HIV/AIDS (30). In this study, we hypothesized and presented data suggesting that the Pneumocystis phosphoglucomutase (PGM) enzyme could be a novel promising target for therapy against Pneumocystis. Previous research has focused on this enzyme in medically important bacteria such as Brucella meltiness, Pseudomonas aeruginosa, and Salmonella typhimurium, demonstrating its significance in virulence (22, 37, 38). Although there has been extensive identification and characterization of PGM activity in S. cerevisiae (15, 30, 39), there has been little work on the importance of the protein in pathogenic fungi. Hu et al. (40) demonstrated that silencing the Ganoderma lucidum [causative agent for root rot in wood (41)] fungal PGM resulted in reduced hyphal growth, increased polysaccharide (IPS) production, and significant reduction in β-1,3 glucan production (40).

In the current study, we demonstrate the presence of a functional PGM homolog in both P. jirovecii and P. murina. To identify both Pneumocystis PGMs, we utilized the S. cerevisiae PGM2 protein sequence and searched the P. jirovecii and P. murina protein databases for potential homologs. We successfully identified a single PGM protein from both species at the protein level.

As Pneumocystis cannot be cultured or transformed in vitro, we employed heterologous expression in pgm2Δ yeast to validate the functionality of both P. jirovecii PjPgm2 and P. murina Pmpgm2. Their expression restored the growth defect of the yeast null mutant on both liquid and solid media containing galactose. Additionally, the inclusion of lithium in solid media with galactose led to inhibited growth compared to the complemented strain grown on solid media with galactose alone, suggesting a direct impact of lithium on both PjPgm2p- and PmPgm2p-related PGM activity.

Moreover, the expression of Pjpgm2 and Pmpgm2 in the yeast pgm2Δ mutant resulted in complete restoration of sedimentation rates, indicating the recovery of normal cell wall phenotype characteristics of the parent strain. Protein lysates of the pgm2Δ mutant expressing Pjpgm2 or Pmpgm2 cDNA resulted in significant restoration of PGM activity in the respective null mutant strains, which was significantly inhibited in the presence of lithium. Next, we conducted a preliminary proof-of-concept experiment in which we incubated P. murina in vivo with lithium and observed a significant decrease in fungal viability, comparable to the anti-Pneumocystis drug pentamidine. Finally, we tested a previously characterized PGM enzyme inhibitor, termed ISFP10, with selective 50-fold selectivity against the A. fumigatus PGM enzyme versus the human ortholog at 2 µM (16) against the PmPgm2p, and noted significant dose inhibition of glucose-6-phosphate incorporation with the inhibitor. These preliminary data provide the exciting possibility of testing the ISFP10 compound in the mouse PCP model for use as a new selective therapeutic anti-Pneumocystis compound.

UDP-Glc is an important intermediate metabolite in yeast involved in several biosynthetic pathways, including the biosynthesis of glycogen and trehalose, protein glycosylation, and the formation of cell wall β-glucans (42–44). Therefore, targeting PGM activity in Pneumocystis may be a viable therapeutic strategy for treating individuals with PJP. It is worth noting that lithium is used in humans for the treatment of bipolar disorder (BD) (45, 46). Although the exact mode of action of lithium in the treatment of BD is not well understood, at the neuronal level, the agent is believed to work by reducing dopamine and glutamate levels while simultaneously increasing inhibitory (GABA) neurotransmission (47). Others have shown that lithium used at therapeutic levels can also affect human Pgm1p activity in vitro (14), but the effects of lithium on Pgm1p activity in vivo are, to the best of our knowledge, unknown and appear to be well tolerated in its use since it was medically reported as a treatment for those with maniac disorder in 1949 (48). Therefore, based on our results showing its inhibition of the growth of the pgm2Δ yeast strain restored with either Pjpgm2 or Pmpgm2 in vivo, as well as its direct inhibition of the respective PGM activity in both proteins, further investigation should determine the potential of lithium as a therapeutic agent against Pneumocystis. Additionally, we have shown that the newly reported and selective PGM inhibitor ISFP10 has significant inhibitor activity against PmPgm2p, possibly resembling lithium while minimizing off-target effects on the human isoform. In this regard, these researchers identified that ISFP10 targets the cysteine 353 (C353) of the AfPGM enzyme, which is also present in both the PjPgm2 (position C351) and the PmPgm2 (position C352) proteins. Notably, this residue is absent in the human and mouse Pgm1 proteins, making it a potential specific target for fungal PGM activity. ISFP10 exhibited an IC₅₀ of greater than 100 µM against human PGM, while a significantly lower IC₅₀ of 2 µM was required for AfPGM, demonstrating preferential inhibition towards AfPGM (16).

In summary, we have genetically and functionally confirmed the presence of active PGM proteins in both P. jirovecii and P. murina and shown that the expression of these proteins can restore normal growth in the S. cerevisiae pgm2Δ mutant. Additionally, we have demonstrated that lithium can directly inhibit both PjPgm2p and PmPgm2p activity. The use of lithium itself, as well as potential greater specific analogs such as the small molecule ISFP10, may be viable therapeutic options for individuals with Pneumocystis pneumonia.

ETHICS APPROVAL

This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals (49). The Mayo Clinic Institutional Animal Care and Use Committee approved this research (A00002337).