Introduction

Pneumocystis pneumonia (PcP) is a severe opportunistic infection that has a high impact on patients with AIDS and is therefore of growing importance among non-HIV-infected subjects who are submitted to immunosuppressive treatments related to malignancies, connective tissue diseases or organ transplantation [1, 2, 3, 4, 5, 6]. The source of PcP infection remains unknown. It is generally accepted that most PcP cases do not result from reactivation of endogenous latent forms, as was previously thought [7, 8, 9], but rather from de novo airborne infection contracted from unidentified exogenous sources [10, 11]. However, PCR assays to detect Pneumocystis spp. have revealed that hospitalized patients without PcP can be latently infected with Pneumocystis spp. [5, 12, 13, 14, 15, 16, 17, 18, 19]. Therefore, they might be able to transmit the organism to susceptible patients via an airborne route.

The hypothesis has practical importance, especially considering that carriers of Pneumocystis spp. in hospitals are usually more numerous than PcP patients [14]. Furthermore, there have been reports of Pneumocystis spp. being detected by molecular methods in immunocompetent contacts (i.e., mother, nurse and physician) of a hospitalized infant with PcP [20] and in immunocompetent healthcare workers who did and did not have direct contact with immunocompromised patients [21, 22]. Infection with Pneumocystis spp. could thus be transmitted in hospitals through healthcare workers. A more recent report suggested that human-to-human PcP transmission could occur from a mother to her infant on the basis of Pneumocystis jirovecii genotyping at three independent DNA loci [23].

No data are currently available about the role of human carriers in PcP transmission, but recent experimental data we obtained suggests they could represent a risk for immunosuppressed patients [24]. In that study, immunocompetent mice transiently parasitized by Pneumocystis carinii forma specialis muris (Pneumocystis muris) after close contact with infected SCID mice did transmit the infection aerially to uninfected SCID mice [24]. Thus, immunocompetent contacts of PcP patients, such as members of hospital staff who have contracted a subclinical Pneumocystis infection from PcP patients, could be able to transmit the PcP agent via the airborne route to severely immunosuppressed patients in hospitals. This possibility is being explored in a long-term study developed at the Lille-2 University Hospital (Lille, France) within the framework of the European project “Eurocarinii” (“Human Pneumocystis carinii infection in immunocompetent hosts: a public health issue in Europe.” Project QLK2-CT-2000-01369).

An additional aim of the present study was to determine whether Pneumocystis organisms are able to circulate aerially among healthy hosts. These hosts are usually able to radically eliminate the parasites from their lungs, as attested by PCR-positive results turning negative [25]. However, as long as the hosts remain infected, they are able to transmit the infection either to naïve immunocompetent hosts, like infants who will develop a primary infection, or to immunosuppressed hosts, who may then develop clinical PcP, thus amplifying the population of Pneumocystis organisms. In this manner, healthy hosts could behave as a sort of dynamic reservoir of Pneumocystis spp. These aspects were recently explored by Gigliotti et al. [26] using the same experimental design used by Dumoulin et al. [24] (i.e., airborne transmission among co-housed SCID and Balb/c mice). Gigliotti et al. [26] showed clearly that PcP infection can pass through normal hosts; in other words, immunocompetent hosts can transmit the infection to other immunocompetent hosts, and that these ‘second’ contacts are able to pass the infection on to susceptible hosts. In the present report, we describe similar results. However, in our experiments, a co-housing period of just 1 day was found to be enough for Balb/c mice to contract the infection from SCID mice with PcP. This differs from the study of Gigliotti et al. [26] which used an invariable co-housing period of 3 weeks.

Indeed, the main objective of the present work was to explore the behavior of Pneumocystis organisms in the lungs of immunocompetent hosts. Specifically, we sought to determine if the parasites carried by these hosts were able to achieve actual cell division in the host’s pulmonary alveoli or if they were viable latent forms able to pass aerially to other hosts. A strategy based on both histopathology and cell-division-cycle 2 (cdc2) gene-targeted reverse transcriptase polymerase chain reaction (RT-PCR) [27, 28] kinetic study was displayed. The findings are important since the demonstration of actual proliferation of Pneumocystis organisms in the lungs of immunocompetent hosts strengthens the recently formulated hypothesis that these hosts act as a reservoir [22, 26, 29].

Materials and Methods

Experimental Design

Two experiments were performed using Balb/c and SCID mice in this study. In both experiments all groups of mice were housed in separate HEPA-filtered air isolators, and drinking water, food and bedding were sterilized.

Experiment 1 was designed to explore the ability of immunocompetent Balb/c mice transiently infected with Pneumocystis muris (carriers) to transmit the infection to other immunocompetent Balb/c mice (contacts). Initially, Pneumocystis muris-free SCID mice (donors) were inoculated intranasally with mouse-derived Pneumocystis. Eight weeks later, Pneumocystis muris-free immunocompetent Balb/c mice were housed in close contact (i.e., in the same cage) with the Pneumocystis muris-infected donor SCID mice for 1 day (Balb/c mice group 1) or 5 days (Balb/c mice group 2) in a proportion of three donor SCID mice to seven contact Balb/c mice.

At the end of the co-housing period the following steps were taken. (i) SCID mice were sacrificed and their lungs were removed for microscopic detection of Pneumocystis muris organisms, PCR detection of Pneumocystis DNA and molecular identification of Pneumocystis muris. (ii) One carrier Balb/c mouse from group 1 and one from group 2 were each co-housed with two new uninfected SCID mice in order to verify that Pneumocystis muris-infected contacts were able to transmit the organism to susceptible hosts. (iii) The six remaining immunocompetent carrier mice from groups 1 and 2 were co-housed respectively with groups 3 and 4, each of which contained six uninfected immunocompetent contact Balb/c mice.

Each of the six mice in groups 1 and 3 were sacrificed sequentially after 1, 4, 6, 8, 12 and 22 days of co-housing. In the same way, mice from groups 2 and 4 were sacrificed sequentially after a co-housing period of 1, 3, 7, 10 and 17 days. The duration of co-housing for individual mice in all groups is presented in Table 1. Additionally, one mouse from group 4 (mouse no. 26, Table 1) was isolated and co-housed with two uninfected SCID mice in order to verify that Pneumocystis muris-infected ‘second’ contacts were able to transmit Pneumocystis infection to susceptible hosts.

Table 1 Transmission of Pneumocystis carinii f. sp. muris (Pc muris) among immunocompetent hosts: morphological and molecular detection of Pc muris in Balb/c mice co-housed either with Pc muris-infected donor SCID mice (for 1 or 5 days) or with immunocompetent Pc muris carrier Balb/c mice (from 1 to 22 days) (Experiment 1, see text)
Full size table

Ten Pneumocystis muris-free SCID or Balb/c mice were housed separately, serving as sentinel controls. They were also sacrificed for detection of Pneumocystis muris at the same time as those from groups 1 to 4.

Experiment 2 was designed to explore the ability of Pneumocystis muris organisms transiently carried by Balb/c mice, to develop actively in their lungs. Ten Pneumocystis muris-free SCID mice (donors) were inoculated intranasally with mouse-derived Pneumocystis. Sixty days later, 18 immunocompetent Balb/c mice that were known to be Pneumocystis muris-free were co-housed with the Pneumocystis muris-infected donor SCID mice for 5 days in a proportion of five donor SCID mice to nine contact Balb/c mice in each cage. Then, carrier Balb/c mice were sacrificed and their lungs were removed sequentially at 0, 5, 9, 12, 15, 18, 21, 25, 30 and 35 days after the end of the co-housing period. Microscopic detection of Pneumocystis organisms, histological analysis [30], extraction and amplification of Pneumocystis DNA or RNA, and molecular identification of Pneumocystis muris were performed on the lungs. Serum was sampled from all Balb/c mice used in this experiment for detection of Pneumocystis-specific antibodies (see below).

Source of Mouse-Derived Pneumocystis Organisms

Mouse-derived strains of Pneumocystis muris were obtained from corticosteroid-treated Balb/c/BU mice (Lille Pasteur Institute, France), and parasite extraction was performed as described previously [31]. Parasite isolates were identified as Pneumocystis muris using molecular methods [32]. Fungal samples of mouse origin were stored in liquid nitrogen until use by means of a cryopreservation protocol used successfully for many years in our laboratory [33]. This method has proved to be efficient for obtaining viable, infectious and ultrastructurally well-preserved Pneumocystis parasites [32, 33, 34].

Source of SCID Mice

Pneumocystis muris-free 7-week-old CB17-SCID mice were used. They were obtained from a colony bred in the Pasteur Institute of Lille, France. All mice were administered dexamethasone 2 mg/ml in drinking water from 15 days pre-inoculation until 8 weeks post-inoculation. It was previously shown that even if SCID mice are basically immunosuppressed hosts, by administering dexamethasone higher and more reproducible parasite rates can be achieved [35]. Microscopy and molecular methods confirmed the absence of latent Pneumocystis infection in the lungs of SCID mice sampled prior to parasite inoculation. SCID mice were anesthetized with an intraperitoneally administered drug cocktail (ketamine hydrochloride 150 mg/kg + diazepam 2.5 mg/kg + atropine 0.75 mg/kg) then inoculated intranasally under sterile conditions with 106 Pneumocystis muris organisms per animal suspended in 25 µl of Dulbecco’s modified Eagle’s culture medium [34].

Source of Balb/c Mice

Immunocompetent Balb/c mice from a selected, known Pneumocystis muris-free colony (Charles River, France) were used as Pneumocystis contacts. By administering dexamethasone (2 mg/ml in drinking water) to 10 mice in the colony and sacrificing them 63 days later for Pneumocystis DNA or microscopic parasite detection, we were able to confirm the absence of latent Pneumocystis infection.

PCR Detection of Pneumocystis DNA

DNA was extracted using the QIAmp DNA mini-kit (QIAGEN, Courtaboeuf, France) processed according to the manufacturer’s recommendations for tissue protocol. PCR detection of Pneumocystis DNA was performed, based on the method of Wakefield et al. [36], at the large subunit of the mitochondrial ribosomal RNA gene using primers pAZ102-H and pAZ102-E. Nested-PCR with the internal primers pAZ102-X and pAZ102-Y was used when the amount of Pneumocystis muris in the sample was low and insufficient to produce a visible band on the ethidium bromide-stained gel after a single round of amplification [37]. The reaction mixtures were prepared in 1×PCR buffer (75 mM Tris HCl, pH 8.8—20 mM [NH4]2SO4—0.01% Tween 20) and contained, per 50 µl reaction, 3 mM of MgCl2, 0.5 µM of both primers (Eurogentec, Seraing, Belgium), 400 µM of each dNTP, 1 U of Goldstar DNA polymerase (Eurogentec) and 20 µl of the purified DNA in the first round of amplification. In the second round, 2 µl of the first PCR product was used, and the PCR conditions were the same as those of the first round. DNA amplification was carried out using a PTC 200 thermocycler (MJ Research, Foster City, CA, USA). The conditions for amplification were 40 cycles of denaturation at 94°C for 1.5 min, annealing at 55°C for 1.5 min and extension at 72°C for 2 min. PCR product was analyzed by electrophoresis in a 2% agarose gel and was visualized after ethidium-bromide staining. Negative controls with no added DNA were included in the DNA extraction procedure and in the amplification run.

Hybridization Method

When analysis of PCR products by direct sequencing was not possible (i.e., no amplification product was visualized on agarose gels), hybridization of PCR products with an oligonucleotide probe specific to Pneumocystis muris (IPL-M) was used as described previously [32].

RNA Extraction

In order to preserve RNA, the lung samples of both Balb/c and SCID mice used in experiment 2 were put into RNA later solution (QIAGEN) immediately after sacrifice. RNA extraction was performed using the RNAeasy Protect midi kit (QIAGEN). Briefly, lung tissues were lysed in adequate buffer and homogenized using the PT3100 system (Kinematica Polytron, Brinkmann, USA). Extraction was then performed according to the manufacturer’s instructions.

RT-PCR on cdc2 Gene

In this study, the primers used for RT-PCR were based on those previously described for the rat-derived-Pneumocystis cdc2 gene [28] and were as follows: CDC2-A (5’-ATG-GAG-CAA-TAT-CAG-AGG-TTA-GAG-3’) and CDC2-B (5’-CTA-TAG-CAC-CAC-ATT-AGA-TCT-ATT-3’). Procedures were based on single-tube RT-PCR using the Titan One Tube RT-PCR System (Roche Diagnostics, Meyran, France). Each 50 µl of reaction mixture contained 5 µl RNA, 0.5 µM of each primer, 0.2 mM of each dNTP, 4 mM of MgCl2, 5 mM of DTT, 5 to 10 U of RNase inhibitor, 5×RT-PCR buffer pH 7.5, and 1 µl of enzyme mix (AMV and Expand High Fidelity PCR-system; Roche Diagnostics).

Reaction mixtures were incubated at 50°C for 45 min and 94°C for 4 min then subjected to amplification for 30 cycles of denaturation at 94°C for 1 min, annealing at 56°C for 1 min and extension at 72°C for 1 min, followed by incubation at 72°C for 10 min. The RT-PCR products were analyzed by electrophoresis in a 1.5% agarose gel and were visualized after ethidium-bromide staining. Negative control reactions included one with RNA from a Pneumocystis muris-free mouse lung, one with a sample that had not undergone reverse transcription, and one from which the nucleic acid template was omitted in order to confirm that the PCR products were not due to contamination. Sensitivity of the cdc2 RT-PCR was evaluated on total RNA samples prepared from log-unit dilutions of Pneumocystis organisms extracted from Pneumocystis muris-infected SCID mice.

RT-PCR on Pneumocystis SA1 Gene

Specific primers of the mRNA PcSA1 gene, member of the Pneumocystis carinii heat shock protein 70 multigene family [38, 39], were used in a RT-PCR to assess organism viability. The Titan One Tube RT-PCR System (Roche Diagnostics) was again used. The reaction mixtures were identical to those described above for cdc2, and the primers or RT-PCR conditions were the same as those used by Maher et al. [38] with some modifications. Briefly, samples were incubated at 50°C for 45 min, followed by 40 cycles at 94°C for 1 min, 50°C for 1 min and 72°C for 1 min, and a final 7-min extension at 72°C. RT-PCR products were electrophoresed on 1.5% agarose gels and revealed with ethidium bromide. The negative controls were the same as those used for cdc2 RT-PCR.

Cloning, DNA Sequencing and DNA Sequence Analysis

Amplified PCR and RT-PCR products were purified by filtration using microcon 50 (Amicon, Beverly, MA, USA). Amplified RT-PCR products were cloned into the pCR2.1 vector using the TOPO TA Cloning kit (Invitrogen, Paris, France). Recombinant plasmids were sequenced in both directions using a model ABI 377 automated sequencer with the Big Dye Terminator Cycle Sequencing kit (Perkin Elmer-Applied Biosystems, Foster City, CA, USA), according to the manufacturer’s instructions. The NCBI BLAST program (www.ncbi.nlm.nih.gov/BLAST/) was used to search for sequence identity in the sequence databases.

Histological Study

Lung samples were fixed in a 10% buffered formalin solution, embedded in paraffin, sectioned (4–5 µm) and stained with hematoxylin-eosin-saffron, Masson trichromic, reticulin and orcein [30, 40]. For the detection of Pneumocystis cystic forms, Grocott’s modified methenamine silver stain was used [40].

Detection of Pneumocystis-Specific Antibodies

A Western blot assay was used to detect Pneumocystis antibody in Balb/c mouse serum diluted 1/50 [8]. The antigens used were cryopreserved Pneumocystis carinii organisms from Wistar rats with steroid-induced pneumocystosis [35]. Anti-Pneumocystis antibodies were detected using a goat anti-mouse immunoglobulin-alkaline phosphatase-labeled antibody (Southern Biotechnology, Birmingham, Ala., USA). Sera from Balb/c mice before co-housing with infected mice were used as negative controls. A monoclonal antibody specific for Pneumocystis muris (Dako, Glostrup, Denmark) was used as positive control.

Results

Pneumocystis Infection in Donor Mice

All SCID mice infected with Pneumocystis muris developed PcP with total parasite rates ranging from 5 to 20×106 organisms per animal.

Pneumocystis Carriage in Balb/c Mice after Co-Housing with Donor SCID Mice

Results of PCR and microscopic examination performed on lung samples from immunocompetent Balb/c contact mice in experiment 1 are shown in Table 1. All lung samples from Balb/c mice co-housed with infected donor SCID mice for 1 or 5 days yielded positive first- or second-round PCR results (groups 1 and 2, Table 1). No mouse from group 1 (1 day of co-housing with SCID mice) showed parasites during the microscopic examination; however, rare Pneumocystis cystic forms were observed in the lung imprint of one mouse (no. 13, Table 1) from group 2 (5 days of co-housing with SCID mice) (Fig. 1).

Fig. 1
figure 1

Pneumocystis cystic forms observed in dried lung imprints of mouse no. 13 (Table 1), which was co-housed for 5 days with Pneumocystis muris-infected SCID mice (donors) and sacrificed 28 days after the end of the co-housing period. Toluidin blue stain. Direct magnification: ×1000

Full size image

In experiment 2, all lung samples from the 18 contact Balb/c mice, collected from the end of the 5-day-co-housing period with the donor SCID mice through the end of the experiment (40 days), showed positive first-round PCR results at a large subunit of the mitochondrial ribosomal RNA locus (Table 2).

Table 2 Pneumocystis carinii f. sp. muris (Pc muris) in immunocompetent Balb/c mice: histological and molecular detection, viability, cell division (see text) and specific serum antibody response (Experiment 2, see text)
Full size table

Pneumocystis Transmission Among Immunocompetent Hosts (Table 1)

Three of the 12 Balb/c mice in groups 3 and 4 showed positive PCR results: one mouse in group 3 (no. 20) after a 22-day-period of co-housing with mice in group 1, and two mice in group 4 (no. 25 and 26) after a 17-day co-housing period with mice in group 2. Rare cystic forms were microscopically observed in the lung imprint of mouse no. 26 (group 4).

Viability and Multiplication of Pneumocystis Organisms in Immunocompetent Carriers

All Balb/c mice in experiment 2 showed positive PcSA1 RT-PCR results, indicating that Pneumocystis organisms carried by these immunocompetent hosts were viable (Table 2). Moreover, three Balb/c mice sacrificed at 0, 5 or 21 days after the 5-day co-housing period with donor SCID mice, were found to express the cdc2 gene, suggesting that parasites were able to achieve cell division in the host lungs (Table 2).

Pneumocystis muris Large Subunit of the Mitochondrial Ribosomal RNA Gene, Heat Shock Protein 70 and cdc2 Gene Amplified Sequences

Using the Pneumocystis-specific oligonucleotide primers pAZ102-E and pAZ102-H at the large subunit of the mitochondrial ribosomal RNA locus, the sequence of amplified products identified Pneumocystis muris in positive lung samples from both the SCID and Balb/c mice from both experiments. For experiment 2, RT-PCR amplified products from PcSa1 (GenBank accession number AY382182) and cdc2 (GenBank accession number AY382183) genes were cloned, sequenced and compared with those from other fungi, humans or other Pneumocystis spp. using the NCBI Blast program. Analysis of the results obtained at the two loci showed the highest homology with rat-derived Pneumocystis. Actually, for the two amplified sequences (cdc2 and PcSa1 gene fragments) an identity of 89% was found between mouse-derived Pneumocystis and rat-derived Pneumocystis. These results allow the conclusion that sequences amplified from SCID or Balb/c mice following RT-PCR of cdc2 or PcSA1 are new Pneumocystis muris cDNA sequences.

Pneumocystis muris in the Lungs of Immunocompetent Hosts

Grocott’s modified methenamine silver stain revealed rare clustered cystic forms, closely associated with lung tissue, lining the alveolar walls of Balb/c mice sacrificed 21, 25, 30 or 35 days after the end of the 5-day co-housing period with donor SCID mice (Table 2, Fig. 2). Associated pulmonary changes were mild and parasite groups were apparently not associated with inflammatory changes. No fibrosis was observed. However, moderate mononuclear infiltrates were noted around bronchial or vascular sections. Vascular congestion and a few alveolar hemorrhagic infiltrates, surely related with post-mortem changes, were also recorded.

Fig. 2
figure 2

Pneumocystis carinii f. sp. muris in the lungs of healthy Balb/c mice. Clustered Pneumocystis cysts (arrows) attached to the alveolar epithelium (a) or placed depth in the alveolar tissue (c). This mouse contracted the Pneumocystis infection by co-housing for 5 days with SCID mice with pneumocystosis (donors). After the co-housing period the mouse was isolated from donors and sacrificed 35 days later. Direct magnification: ×160 (a) and ×200 (c). In micrographs b and d, parasitic clusters (arrows) are clearly visible closely attached to the alveolar epithelium of a healthy Balb/c mouse. It had been co-housed for 5 days with Pneumocystis-infected donor SCID mice, and then isolated from them for 21 days. Grocott’s methenamine silver stain. Direct magnification: ×1000 (b and d)

Full size image

Pneumocystis-Specific Immunoglobulin in the Sera of Immunocompetent Carriers

Pneumocystis antibody was detected in the sera of the Balb/c mice as early as 17  days after the start of the 5-day co-housing period with donor mice and henceforth (Table 2).

Sentinel Mice

Ten sentinel mice (4 SCID mice and 6 Balb/c mice) sacrificed during the experiment were found to be negative by both microscopy and nested-PCR assay.

Transmission of Pneumocystis muris to SCID Mice Through Immunocompetent Balb/c Mice

Pneumocystis muris DNA was detected by PCR in all of the SCID mice (6/6) co-housed for 45 days with either the contact mice of groups 1 and 2 (no. 7 and 14, Table 1) or with one contact mouse of group 4 (no. 26, Table 1). Parasites were observed on the lung imprint of all SCID mice.

Discussion

The understanding of Pneumocystis infection mechanisms is a crucial epidemiological challenge, since the source and reservoir of the infection in humans have not yet been established. The infection can be transmitted aerially [24, 41, 42], and cross-infection experiments have shown that a Pneumocystis sp. from a given mammal cannot infect other host species [32, 34, 43, 44]. Pneumocystis jirovecii could behave in a similar manner, and humans could therefore constitute the reservoir for this species. Pneumocystis DNA has been detected consistently in hospitalized patients without PcP and in immunocompetent hospital staff members who either had or did not have contact with immunocompromised patients [20, 21, 22].

The detection of Pneumocystis carriers among immunocompetent healthcare workers led us to wonder about their role as a possible source of infection for either susceptible inpatients or other healthcare workers. Thus, a recent study was conducted by our team that demonstrated that Pneumocystis muris can be transmitted from SCID mice with PcP to immunocompetent mice, which were then able to transmit the pathogen to susceptible hosts [24]. Only 1 day of contact was enough to transmit the infection [24, 42]. In the present study the period examined was 1 or 5 days, which seems more representative of clinical situations in a nosocomial context than the 3-week co-housing period used by Gigliotti et al. [26] in a recent study.

Regarding the transmission of Pneumocystis organisms among immunocompetent hosts, the present study confirmed that immunocompetent mice transiently parasitized by Pneumocystis spp. were able to transmit the organism to other immunocompetent Pneumocystis-free mice. However, healthy host-to-healthy host transmission was not detected in all cases (3 of 12 healthy mice) and a minimal co-housing period of 17 days was required for transmission in our conditions (Table 1). In addition, we demonstrated that these ‘second’ healthy contacts were able to transmit the infection to susceptible hosts, a result that is consistent with the recent observations of Gigliotti et al. [26]. Indeed, the results of the present work suggest that even a weak parasite load is enough to transmit the infection from healthy hosts to susceptible hosts, as in the case of mouse no. 26, which contracted the infection from a healthy carrier (Table 1).

On the whole, our results show the potential importance of immunocompetent carriers in the circulation of Pneumocystis spp. in host populations. A further aim of our study was to explore the potential role of immunocompetent hosts as a reservoir for Pneumocystis organisms. Gigliotti et al. [26] showed that the infection passes through normal Balb/c mice and, from this fact, they concluded that active replication takes place in immunocompetent hosts. However, transmission may also occur by airborne transfer of latent forms of Pneumocystis spp. without increasing their population. This is a key point because only the demonstration of actual parasite replication in the lungs of immunocompetent hosts allows them to be considered reservoirs for Pneumocystis spp. For this reason, we attempted to demonstrate actual cell division of Pneumocystis in the lungs of healthy hosts.

In this part of our study, we detected Pneumocystis cdc2 mRNA in the lungs of immunocompetent Balb/c mice that had contracted a transient Pneumocystis infection by a natural airborne route from Pneumocystis muris-infected SCID mice. The cdc2 gene encodes a cyclin-dependent serine-threonine kinase whose activity is essential for cell division [45, 46, 47]. The presence of this gene has been detected previously in the genome of Pneumocystis carinii by Thomas et al. [28] and in the genome of Pneumocystis jirovecii by Kaiser et al. [48, 49], who revealed in vitro that the expression of this gene is correlated with the growth rate of the organism [48]. We successfully detected the mouse-derived Pneumocystis cdc2 gene in Balb/c mice sacrificed 5, 10 and 26 days after the start of the 5-day co-housing period, suggesting that active replication of Pneumocystis organisms occurred in these immunocompetent hosts.

The RNA preservation or extraction methods we used were validated by the fact that PcSA1 mRNA was detected in all of the lung samples of Balb/c contact mice. RT-PCR amplification of the mRNA of a member of the heat shock protein 70 family, which is usually employed to assess viability, has been reported previously for rat and human-derived Pneumocystis spp. [38, 39]. The method was successfully applied in our study enabling us to amplify for the first time the homologous mouse-derived Pneumocystis heat shock protein 70 mRNA fragment.

The histological finding of parasites in Balb/c mice sacrificed 26 to 40 days after the 5-day co-housing period provides a morphological basis for the observations resulting from the molecular approach and strengthen their significance. The presence of clustered parasites (Fig. 2) closely attached to the alveolar cells suggests that they actively replicate in this location. Further evidence of the close interaction between Pneumocystis muris and the immunocompetent host was provided by the detection of antibodies to Pneumocystis in the sera of the Balb/c contact mice, suggesting that seroconversion took place. The first detectable serum antibody response, revealed by using an anti-mouse immunoglobulin-alkaline phosphatase-labeled antibody, was observed as early as 17 days after the start of the co-housing period (Table 2). This result is consistent with a recent report describing the detection of Pneumocystis muris-specific immunoglobulin G in the sera of immunocompetent mice after about 4 weeks of co-housing with SCID mice [50].

In conclusion, our study has demonstrated that healthy hosts transiently parasitized by Pneumocystis organisms could serve as a reservoir for Pneumocystis spp. Extrapolation of these results to the natural history of human PcP leads us to hypothesize that PcP patients, immunodeficient carriers (or colonized patients) and transiently parasitized immunocompetent individuals can all play a role as both a source of infection and a reservoir of Pneumocystis jirovecii. These new data suggest that infection with Pneumocystis spp. can be transmitted in a nosocomial pattern. Prevention strategies targeting Pneumocystis carriers should therefore be defined in order to prevent the circulation of Pneumocystis spp. in the hospital environment.