Introduction

Prototheca species are unicellular, achlorophyllous, yeast-like microalgae thriving in a wide range of environmental habitats, typically with high moisture and organic matter content [1]. These organisms normally live a saprophytic lifestyle, but can become opportunistic pathogens and cause a variety of pathologies in both animals and humans, collectively referred to as protothecosis [1]. There was a century-long debate over the taxonomic status of Prototheca until it reached a final consensus with their placement in the Trebouxiophyceae class within the Chlorophyta lineage. Prototheca have been accepted to represent descendants of green algae of the genus Chlorella, which, during their evolutionary history, have lost the ability to photosynthesise and changed to a heterotrophic metabolism [1,2]. The current taxonomic classification of Prototheca has been built on the mitochondrially-encoded cytb gene, with a total of 18 distinct species described [3,4]. Five of these species have been implicated in animal protothecosis (P. blaschkeae, P. bovis, P. ciferrii, P. cutis, P. wickerhamii) [5–7], whereas the same five species and P. miyajii have been involved in human infections [8–10].

The most prevalent form of animal protothecosis is bovine mastitis, usually recognized as a chronic, subclinical infection, for which culling is the only control strategy, since no effective treatments exist. The disease has a global distribution and incurs heavy economic losses to the dairy industry and animal welfare [7,11–13]. Contrastingly, human protothecosis is a rare yet an emerging infection, whose incidence has been on the rise over the last few decades [8,14]. Typically, it manifests under three major clinical forms, namely cutaneous, articular (olecranon bursitis), and disseminated (or systemic) infections. Patients suffering from different types of immunosuppression are most commonly affected [8,10]. Of particular concern are systemic infections, for which the mortality rate is exceptionally high, exceeding 50% on average [14]. A hallmark feature of Prototheca infections is their refractoriness to most of the therapeutic regimens currently available with virtually no mechanisms behind this resistance disclosed [8,10,14,15].

Although Prototheca were first recognized as pathogens in the early 1950s [16], rarely have they been studied scientifically. Therefore, there is a huge knowledge gap on Prototheca pathobiology and virulence mechanisms. To the best of authors’ assessment, only slightly more than 20 studies have addressed the problem of Prototheca pathogenicity with the use of animal models [17–40]. In five of these studies, animal challenge with the algal culture was part of the diagnostic path described in case reports [35–39]. All in vivo studies have been reviewed and summarized in Table 1. Among several animal species experimentally inoculated with Prototheca algae were rabbits [19,22,37], guinea pigs [19,22,23,35,37], rats [22,37,40], rhesus monkey [40], mice [17,18,20–23,25–30,32–34,36–40], and dairy cows [20,24,31]. Recently, a murine model for studying Prototheca bovine mastitis has been proposed [32]. The inoculum dose required to cause an infection, differed across studies, ranging from 40 to 4.5 × 10⁸ cells [20,24]. The most frequently exploited route of inoculation was the intraperitoneal route (I.P.) [19–23,25,30,32,36,37,39], followed by subcutaneous (S.C.) [19,20,22,26,28,29,32,37], intramammary (I.M.) [17,18,20,24,31,32,34], intratesticular (I.T.) [19,22,25,38], intravenous (I.V.) [19,22,25,27], and intradermal (I.D.) [23,28,29,35] exposure. Of these, I.M. and I.T. routes proved most efficacious, invariably leading to an infection. The I.D., I.P., I.V., S.C. infection routes were somewhat less effective, but still resulted in infection in at least half of the animals challenged [19–21,25–30,32,36,39]. In single cases, alternative inoculation methods were employed, such as intramuscular (I.MS.) [20], intratracheal (I.TR.) [21], oral (P.O.) [40], and also topical (T.) [33], transdermal (T.D.) [32], and intraocular (I.O.) [37] delivery, each but the latter two producing an infection.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Studies using animals in experimental protothecosis – literature review.

https://doi.org/10.1371/journal.pntd.0014312.t001

Of all in vivo studies thus far performed, only four have investigated the pathogenic potential of Prototheca algae in a comparative manner, that is using two or more Prototheca species [20,22,30,40]. One such study showed that P. zopfii (currently P. bovis or P. ciferrii) was twice as successful in causing infections compared to P. wickerhamii [30]. All but four in vivo studies employed immunocompetent animals. In one such investigation, neutropenic guinea pigs and athymic mice were used [23], whereas the remaining three involved mice that were compromised by exposure to either prednisolone [25,27] or dexamethasone [33].

In rare instances where replicable animal models, infection doses, and routes were used, the results varied considerably between the studies [20,22,27,32,37]. The discrepancies in the findings on the pathogenic capacity of Prototheca spp. largely arise from the variability in experimental approaches. Earlier in vivo studies differed significantly in their experimental parameters, including the algal strain, infection dose, infection route, host species, and the spectrum of tissues examined. This methodological variability has led to inconsistent and often irreproducible results, making meaningful comparisons across studies challenging or even unfeasible. Moreover, the very narrow scale of the previous investigations has further hampered gaining a coherent view of the Prototheca infectivity potential.

The only consistent pattern that emerged throughout most of the studies was that the frequency of infection and host mortality conspicuously increased with higher inoculum loads [21,27,32]. Still, the limitations in the design of previous studies preclude determination of the impact of other experimental parameters, including the algal species and the host species on the infection development and severity.

The present study was conceived to address these limitations through launching, for the first time, a large-scale, multivariable investigation, involving different types of pathogens, host species, inocula, and routes of infection. The purpose of the study was to elucidate the role of these essential experimental components on the induction and extent of Prototheca infection in the in vivo model.

Materials and methods

Mice

The study sample included a total of 324 female, 8-week-old mice, split into two cohorts of equal size (162 subjects), each representing a different genotype. The first cohort included wild-type Balb/cAnNRj mice, referred to as wild-type (wt), whereas the second cohort consisted of immunodeficient mice (Balb/cAnN-Foxn1nu/nu/Rj) referred to as nude or athymic mice.

All mice were purchased from Janvier Labs (France), delivered to and housed in the Central Laboratory of Experimental Animals of the Medical University of Warsaw. Once arrived, the animals were quarantined for at least 1 week before use to adapt to the new housing conditions. The mice were kept in groups of six animals (i.e., each experimental group separately) in transparent Sealsafe Plus GM500 IVC cages (Tecniplast, Italy) providing a minimum area of 0.05 m². Rooms where cages were placed were equipped with mechanical ventilation (15 changes per hour), regulated daylight (12 hours day/12 hours night), temperature ranging from 20°C ± 4°C, and humidity maintained at 55 ± 10%. Food (Altromin 1318 feed (Animalab, Poland)) and water were available ad libitum. Cages, feed, water, and enrichment materials were sterilized by autoclaving.

Mice of each strain were split evenly into 27 groups of 6 subjects per group (54 groups in total), depending on the inoculum (algal strain or PBS as a control), the challenging dose (a 0.2-mL aliquot of the algae containing 5x10⁶ or 5x10⁷ CFU/mL, or an equal volume of PBS), and the inoculation route (subcutaneous, intramammary, and intraperitoneal). The inocula were injected using one-milliliter syringes with a replaceable needle to either subcutaneous tissue in the neck area, the 4^th left mammary gland, or intraperitoneally in the lower abdominal area. A schematic diagram of the experimental path is presented in Fig 1.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Experimental path in the murine model of Prototheca infection conceived in this study.

https://doi.org/10.1371/journal.pntd.0014312.g001

Examination and sampling

Over a 6-week experimental period, the animals were observed for behavioral and anatomical changes. Their weights were recorded twice, i.e., on the day of inoculation and on the terminal day. Humane endpoints and criteria for earlier euthanasia were defined solely as precautionary measures and were not reached during the experiment. This included body weight loss exceeding 20%, markedly reduced activity, significantly decreased food intake, injection-related complications causing pain or distress, or other severe adverse signs such as bleeding, pallor of the skin or mucous membranes, or respiratory difficulties.

Six weeks post-inoculation, mice were first anaesthetized by inhalation of isoflurane and euthanized by dislocation of the cervical vertebrae. Immediately after euthanasia blood was collected directly from the heart using one-milliliter disposable syringes, and transferred to the test tube containing EDTA as an anticoagulant. Organs were explanted using sterile instruments in the following order: brain, liver (2 lobes), kidneys, spleen, skin (2 cm²), and mammary glands. If applicable, additional lesions (e.g., abscesses) were also removed. The collected organs were weighed and divided into two portions; one was fixed in 10% formalin, while the other immersed in PBS (pH = 7.4) solution, and sent for histopathological and microbiological examination, respectively. Every sample collected was immediately placed on ice until further analysis. During post-mortem examination, macroscopic changes were systematically assessed, with particular focus on the six major organs analyzed in the study (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Macroscopic and histopathology-based microscopic changes upon Prototheca sp. infection in the brain (A), liver (B), skin (C), kidneys (D), spleen (E), mammary glands (F), and abscesses at the injection site (G).

https://doi.org/10.1371/journal.pntd.0014312.g002

Interpretation and comparison of diagnostic results

Prototheca infection was considered to be induced in a mouse when algae were detected either on microbiology or histopathology evaluation. Infection rate (IR) was defined as the proportion of inoculated animals that developed Prototheca infection. Apart from six internal organs that were routinely examined (i.e., brain, kidneys, liver, mammary glands, skin, and spleen), purulent lesions (typically abscesses at the injection site) were also recorded upon necropsy. These lesions were tested only microbiologically, as they could not be divided. Thus, they were excluded from the analysis comparing the diagnostic performance of histopathology and microbiological examination. Sensitivity of each method was calculated as the proportion of true-positive cases in a given method out of all positives in both methods. A graphical comparison of detection effectiveness and the corresponding IRs are presented in Fig 3.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Distribution of organs affected in individual mice, of either wt or nude phenotype, upon Prototheca sp. infection.

https://doi.org/10.1371/journal.pntd.0014312.g003

Results

Experimental infection

Prototheca experimental infection could be established in nearly a third (29.9%; 43/144) of wt mice and almost a half (45.1%; 65/144) of nude mice (p < 0.01). These animals showed the presence of algae in at least one of the tissues examined, as evidenced upon histopathology or microbiological examination (Fig 3 and Table 2). The infection efficiency varied significantly, depending on the Prototheca species used for inoculation (p < 0.001). Overall, P. ciferrii accounted for the highest infection rate (44/72; 61.1%), followed by P. bovis (33/72; 45.8%), and P. wickerhamii (23/72; 31.9%). Only a small proportion (8/72; 11.1%) of inoculations with P. stagnora produced infection. This species ranking was the same for the two (wt and nude) mouse cohorts analyzed separately. Marked differences were observed between Prototheca pathogenic species (P. ciferrii, P. bovis and P. wickerhamii) whenever wt mice were challenged (IR of 63.9 vs. 33.3 vs. 11.1%, respectively; p < 0.001). This was not reproduced among nude mice, for which the IRs with all the three species were very similar (IR of 58.3% for P. ciferrii and P. bovis, and 52.8% for P. wickerhamii; Fig 4 and Table 2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Distribution of infection in wt and nude mice according to Prototheca species and route of inoculation.

https://doi.org/10.1371/journal.pntd.0014312.t002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Number of mice with and without developed infection following inoculation with different Prototheca species, categorized by mouse phenotype (wt or nude).

https://doi.org/10.1371/journal.pntd.0014312.g004

The organs most commonly affected, and at a comparable rate among wt and nude mice, were mammary glands, followed by skin, spleen, kidneys, and liver. The only significant difference between the two mouse strains concerned brain infections, which occurred five times more often in nude mice compared to wt animals (10.4% vs. 2.1%; p < 0.01; Table 2). Apart from major organs, abscess formation at the injection site was recorded in more than one in ten mice (wt: 11.8%; nude: 17.4%; p > 0.05). In wt animals, the abscesses occurred, at a similar frequency, as isolated lesions or combined with other organ lesions (5.6% vs. 6.3%), whereas in nude mice the suppurative lesions were more often accompanied by other tissue abnormalities (12.5% vs. 4.9%; p < 0.05).

Whereas the mouse strain did not significantly affect the frequency of individual lesions, it did influence the extent of their anatomical dissemination. Multi-organ involvement was clearly more common in nude mice compared to wt mice (15/144; 10.4% vs. 31/144; 21.5%; p < 0.05; Fig 3 and Table 2).

The site of infection was primarily affected by two parameters, namely the algal species and the challenging dose (p < 0.001; Table 2). Among the species tested, P. ciferrii most frequently caused infections of all tissues analyzed, including kidney (75%), spleen (62.5%), skin (57.9%), abscesses (54.8%), liver (53.3%), brain (38.9%), and mammary glands (38.6%; p < 0.001). The rate of infection due to P. bovis and P. wickerhamii differed across the organs, but never exceeded 30% (Table 2). Finally, P. stagnora-induced infections occurred only sporadically and involved either skin (13.2%) or mammary glands (5.3%), with no internal organ involvement (Table 2). Multi-organ involvement was not species-dependent. However, in wt mice, only P. ciferrii was able to produce disseminated infection (Fig 3 and Table 2).

In turn, the frequency and dissemination of infection were strongly affected by the infectious dose (p < 0.001). The higher dose (10⁷ cells) led to considerably elevated infection rates in both wt (25/144; 17.4% vs. 18/144; 12.5%) and nude mice (52/144; 36.1% vs. 13/144; 9%) (Fig 3). Likewise, multi-organ involvement increased markedly when the higher dose was used (42/144; 29.2% vs. 4/144; 2.8%) (Fig 3).

Irrespective of algal species, challenging dose, or host strain, the IR did not differ significantly in terms of inoculation route. The IRs after I.P., I.M. and S.C. injection of Prototheca fell within the ranges of 8.3-58.3%, 9.7-58.3% and 3.4-66.7%, respectively. An exception was P. stagnora, which failed to induce infection via the S.C. route, while I.M. and I.P. inoculations yielded IRs of 20.8% and 12.5%, accordingly (Fig 3 and Table 2). Although the inoculation route did not significantly affect lesion distribution, infections most often involved tissues adjacent to the injection site, with 74.1% (80/108) of cases confined to this region. Thus, with I.M. and S.C. inoculation, the lesions were predominantly found in the mammary gland (59.6%), and the skin (47.4%), respectively. Whereas purulent lesions invariably developed at the injection site (Table 2). Likewise, the occurrence of multi-organ infections did not differ substantially between the inoculation routes, accounting for 17.7%, 13.5%, and 16.7% of cases following I.P., I.M., and S.C. administration, respectively.

Microbiology

The Prototheca infection load in the explanted tissues varied among experimental groups according to three parameters, i.e., Prototheca species, infection dose, and mouse strain. However, no consistent associations with these variables were found (Fig 5). The organs with the highest algal yield were mammary glands. Here, the mean infection load was calculated at 1.4 × 10⁴ CFU/g and 6.5 × 10⁵ CFU/g in wt and nude mice, respectively. Somewhat lower values were observed for the spleen (wt, 1.9 × 10³ CFU/g; nude, 2.2 × 10⁵ CFU/g) and skin (wt, 1.5 × 10³ CFU/g; nude, 5 × 10³ CFU/g). The brain, kidneys, and liver showed the lowest mean infection loads, ranging from 7.2 × 10 to 7.9x10² CFU/g in wt mice, and from 2.6 × 10² to 9.5 × 10² CFU/g in nude mice (p > 0.05; Fig 5).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Algal load in tissues (CFU/g) of wt and nude mice categorized by inoculation dose (A), Prototheca species used for inoculation (B), and tissue type (C).

https://doi.org/10.1371/journal.pntd.0014312.g005

Histopathology

The frequency and the depth of tissue changes correlated significantly with the algal species (p < 0.0001). In general, P. ciferrii and P. bovis induced more pronounced lesions (Table 3). The highest RISI values were recorded for kidneys (6.8-26.1%), skin (0.5-43.5%), and spleen (5-37.1%). Lower values were noted for the mammary gland (1.6-14.6%), liver (0-9.9%), and brain (0–5%) (Table 3). For most tissues, RISI values did not differ when calculated for tissues positive and negative for Prototheca. Exceptions concerned the liver and skin of wt mice, where more intense inflammatory changes were noticed after infection with P. ciferrii and P. stagnora, respectively (p < 0.05). In nude mice, higher RISI values were observed in P. wickerhamii-infected mammary glands, and Prototheca-free samples of the skin and spleen following P. bovis inoculation (p < 0.05; Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Relative Inflammatory Severity Index (RISI) calculated for six major organs, explanted from mice infected with Prototheca species.

https://doi.org/10.1371/journal.pntd.0014312.t003

Blood morphology

Analysis of blood morphology revealed differences associated with mouse strain but not with algal species, infection dose, or route of administration. The levels of WBC, EOS, LYM, and HGB were significantly higher in wt mice compared to nude mice (p < 0.05; Table 4). In turn, MCH and MCV were higher in the latter animals (p < 0.0001). Apart from a slight reduction in MCV in wt mice (p < 0.05), Prototheca infection did not affect hematological parameters, including WBC, LYM, NEU, MONO, EOS, and BAS. Noteworthy, a consistent decrease in HGB and leukocyte counts (including WBC, LYM, EOS, and BAS) was observed in all experimental animals when compared to the reference ranges provided by the breeding facility (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Blood morphology of Prototheca-infected mice.

https://doi.org/10.1371/journal.pntd.0014312.t004

Cytokine profiling

The cytokine profile differed markedly between mouse strains, with wt animals consistently exhibiting higher concentrations of both proinflammatory and regulatory cytokines (p < 0.05; Fig 6), regardless of other infection parameters (algal species, inoculum dose, route of administration). The most pronounced differences were observed for IL-10 (334 pg/mL vs.1.6 pg/mL), TNF (221 vs. 26 pg/mL), and IL-6 (67 vs. 10 pg/mL). Other cytokines, including IL-2, IL-17A, IFN-γ, and IL-4, were likewise elevated in wt mice, although to a lesser extent (22.3-55 pg/mL vs. 0.2-2.8 pg/mL).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Levels of seven different cytokines (A-G) measured in blood samples collected from wt and nude mice at euthanasia, categorized by the Prototheca species used for infection.

https://doi.org/10.1371/journal.pntd.0014312.g006

Of the four Prototheca species tested, the most marked cytokine response was elicited by P. stagnora, albeit only in wt mice (80–846 pg/mL; Fig 6). In contrast, cytokine levels in nude mice infected with this algal species remained comparable to controls (0.2-31.4 pg/mL vs. 0-18.8 pg/mL). The remaining Prototheca species elicited overall weak cytokine responses, with their levels in wt mice not exceeding 50 pg/mL (3.1-48.1 pg/mL) and comparably low concentrations observed in nude mice (0-58.8 pg/mL; Fig 6). No evident differences in cytokine levels were observed in any of the experimental groups, as far as the route of inoculation, infectious dose, or disease progression were concerned.

Discussion

Studying the pathogenicity of Prototheca algae in vivo is very challenging. This is due to the lack of a reliable and validated animal model and the still rudimentary understanding of Prototheca biology. A limited number of previous investigations have produced variable and inconclusive results [17–40]. The present study is the first attempt to develop a comprehensive and robust animal model of protothecosis, whose experimental path was established using a multifactorial design, with a multi-cohort study sample. For the first time, the effects of different algal species, inoculation dose, and route of inoculation, as well as the immunological background of the host, on the development and severity of Prototheca infection in a murine model were systematically explored.

In this study, all Prototheca species examined were able to produce infection in mice, yet the infection potential differed between the species, depending on the experimental conditions. The overall infection rate was calculated at 37.5%, with the highest rate for P. ciferrii (61.1%), followed by P. bovis (45.8%), and P. wickerhamii (31.9%). Whereas, P. stagnora caused infections merely in single animals (11.1%). The pathogenic capacity of Prototheca species was reflected not only in the frequency of infections but also in the range of tissues affected. P. ciferrii produced infections in all organs examined, demonstrating broad tissue tropism. P. bovis was likewise recovered from various organs, albeit the frequency and extent of the lesions were somewhat lower than for P. ciferrii. Contrastingly, infections caused by P. wickerhamii were generally mild and mostly confined to the mammary gland tissue. The few infections due to P. stagnora were of limited severity and were typically localized at peripheral sites, such as the skin or mammary glands.

All four Prototheca species evaluated in this study had previously been shown to exhibit pathogenic potential, although these earlier findings have originated from only a handful of anecdotal studies (Table 1). For P. stagnora in particular, the evidence comes from a single study, which involved intratesticular inoculation of various host species [22]. Only four in vivo studies have directly assessed the pathogenic capacity of different Prototheca species [20,22,30,40], yet one study alone has identified a clear interspecies distinction, with P. zopfii (currently recognized as P. bovis or P. ciferrii) causing about 1.7 times more infections than P. wickerhamii [30].

Differences in virulence between Prototheca species may relate to variation in metabolic plasticity and mechanisms of intracellular persistence. A comparative genomic and transcriptomic study indicated that highly pathogenic P. bovis strains show enhanced expression of genes involved in adaptation to phagolysosomal stress and survival within macrophages [41]. Likewise, previous studies on P. wickerhamii have identified substantial strain-level differences in gene expression [42,43], including genes involved in cell wall biosynthesis and remodeling, such as those encoding mannan endo-1,4-β-mannosidase. Downregulation of this enzyme resulted in thinning of the cell wall and reduced cytotoxicity toward macrophages [43]. This, in turn, may contribute to altered interactions with host immune cells, potentially affecting susceptibility to immune clearance. Such variability in virulence-associated characteristics may also underlie the differences in infection severity observed between species in our study. However, further comparative multiomic studies are necessary to provide a definitive explanation for interspecies variation in virulence.

In this work, a higher inoculum dose (10⁷) was associated with an increased infection rate. This correlation was particularly evident for P. ciferrii in both mouse phenotypes. P. bovis and P. wickerhamii followed the same dose-dependent pattern, but only in nude mice. In contrast, infection rates with P. stagnora remained low regardless of the inoculum size (Fig 3). The trend, linking a larger inoculum with a higher infection rate, had previously been reported, albeit in only one study, where intravenous inoculation of mice with P. zopfii resulted in fewer infections when the inoculum was reduced from 10⁵-10⁷ to 10³-10⁴ [27].

As for the inoculation route, I.M. and I.P. applications were markedly more effective than S.C. injection. This was observed across all tested Prototheca species and mouse strains (Fig 3 and Table 2). Similar to our findings, I.M. inoculation proved to be the most effective route for infecting mice in previous investigations [17,18,32,34] Meanwhile, the I.P. route resulted in infection in over half of the challenged animals, as demonstrated across earlier investigations [20–22,25,30,32,36,39].

In the present study, the infections remained largely subclinical, and only a small subset (2%) of nude mice developed visible lesions, such as skin erosions or nodules at the injection site, after I.P. or S.C. administration. No mortality was reported, and the overall clinical condition of both wt and athymic mice remained stable throughout the whole observation period. Single-organ infections occurred at relatively similar rates in both wt and nude mice, hovering around 20% in both groups. Multi-focal infections were slightly less common, affecting every tenth wt mouse and every fifth nude mouse (Table 2). The frequency of such systemic infections was 10-fold greater when the higher inoculation dose was applied. Neither the inoculation route nor the Prototheca species affected the dissemination pattern of infection.

As shown in earlier research, clinical signs were generally absent, although one study noted a dose-dependent increase in localized redness and swelling at the injection site [32]. Lethal outcomes were reported in at least twelve animals across four investigations irrespective of whether or not the disease had disseminated [19,21,27,40]. Only a single study attributed mortality to I.V. administration and a high inoculum dose (10⁷) [27]. A few studies documented disseminated Prototheca infections in an animal model. These infections concerned rabbits, cows, and mice, exposed to inocula of 10³-10⁷ P. ciferrii, P. wickerhamii or P. zopfii given via I.P., I.T. or I.V injection [19,25,27,30,36].

Algal loads in our experiments averaged from 7.2 to 6.5 × 10⁵ CFU/g in different organs, with the highest mean count observed for the mammary glands and the lowest for the liver. In contrast, former animal studies reported considerably higher tissue burdens, ranging from 1 × 10⁴ to 5 × 10⁵ CFU/g of infected tissue [17,18,32,34]. Interestingly, some studies comparing tissue colonization in mice sacrificed at different post-infection times showed a gradual reduction in algal load over time, a trend consistently observed across different Prototheca species and under various experimental conditions [22,32].

As with other opportunistic pathogens, vulnerability to Prototheca infection and severity of the disease has usually been considered as a function of the immunological status of the host. Among the factors predisposing the development of protothecosis, corticosteroid-induced immunosuppression, neutrophil dysfunction, and impaired T-cell based immunity have most frequently been speculated [23,44,45]. On the other hand, the relatively low prevalence of the disease in HIV-positive and AIDS patients has cast doubt over the role of T cells as critical drivers of the pathogenesis of protothecosis [8,46]. The few animal-based studies evaluating the immune competence of the host have yielded ambiguous and inconclusive results, as some showed higher infection rates in immunocompromised animals versus healthy ones, while the others found the reverse [23,25,27,33].

Although there is no consensus on whether Prototheca algae are intracellular or extracellular pathogens, the literature provides several lines of evidence that the Prototheca lifestyle in the host involves intracellular residence and proliferation. Prototheca have been demonstrated to survive and replicate within phagocytic compartments of macrophages and neutrophils [44,47,48]. The algae have also been shown to invade and persist inside epithelial cells [17,49]. Furthermore, the gene expression profile for pathogenic Prototheca strains differed from that of environmental strains, reflecting a response adapted to the intracellular milieu [41]. From these observations, it was conjectured that a T cell-mediated response might be primarily involved in Prototheca infection. Consistent with this hypothesis, athymic mice, deficient in T lymphocytes, were employed in this study.

Compared to wt mice, the immunosuppressed animals were 1.5 times more susceptible to Prototheca infection with an overall infection rate of 45.1%. Likewise, disseminated infections in nude mice occurred twice as often as in the control animals. These observations appear to argue in favor of the importance of T cell-mediated immunity. Still, half of the athymic mice did not develop infection, which may allude to the role of innate immune response, including phagocytosis by neutrophils and macrophages, and the complement pathway, particularly in the early stages of infection. In this context, it is worth noting that in vitro studies on Prototheca infection have yielded somewhat conflicting results. While some have demonstrated that human and murine phagocytes can internalize Prototheca cells, with this process being species-dependent and often enhanced by IgG-mediated opsonization [17,48,50], the others have indicated that bovine and murine phagocytes may fail to effectively control algal proliferation [44,51]. The scenario is further complicated by the fact that the susceptibility to Prototheca infection and the defective response to Prototheca antigens in nude mice, as a result of lymphocytic depletion, can be mitigated by the activity of still-functional B cells and a relatively normal IgM response to thymus-independent antigens. Thus, to better assess the capacities of isolated cell populations for rendering protection against Prototheca infection, studies employing severe combined immunodeficiency (SCID) mice, lacking both T and B lymphocytes as models, should be considered. Whereas, mouse strains with neutrophil or natural killer (NK) cell deficiencies may provide useful models for clarifying the role of non-specific immunity components [52–54].

Previous studies, both in vitro and in vivo, have demonstrated elevated cytokine levels following Prototheca infection, typically peaking within 12–24 hours post exposure [17,18,34]. The cytokines whose secretion increased consistently in response to Prototheca infection included TNF-α and IL-10 [17,18,34]. These were also found elevated in the present study, with higher values in wt than athymic mice. This difference may indicate a substantial contribution of T lymphocytes in wt mice to cytokine regulation, since both TNF-α and IL-10 can be secreted by activated T cells in addition to macrophages. Sustained TNF-α expression may contribute to infection control by promoting macrophage activation and inflammatory signaling, whereas IL-10 counteracts these effects by limiting macrophage activation and antigen presentation [55]. Interestingly, IL-10 is also a hallmark product of regulatory T cells, which maintain immune homeostasis through ongoing activity in response to commensal microbiota and endogenous antigens [56,57]. Thus, the elevated IL-10 observed in response to P. stagnora may reflect a regulatory rather than pro-inflammatory immune function, consistent with the normally saprophytic (non-pathogenic) phenotype of this species.

Other immune mediators, including IL‑2, IL‑4, IL‑6, and IL‑17A, showed only minor increases. Captivatingly, wt mice infected with the saprophytic P. stagnora showed conspicuously higher cytokine responses than those challenged with pathogenic Prototheca species, suggesting that other factors are involved in shaping the response to the pathogen and channeling the cascade of the inflammatory events and other pathological processes at the cellular level. Such factors, which may differ between species, could include, cell-surface-associated antigens. This speculation is supported by findings in Candida albicans, where cell wall remodeling has been shown to attenuate immune detection and cytokine induction [58,59].

Finally, a comment should be made in regard to the identification of protothecosis in an animal model. Canonically, the diagnostic path for Prototheca spp. includes histopathological examination and tissue culture, both of which were evaluated in this study. Less than half of the mice that had developed Prototheca infection were positive upon histopathology and microbiology, while the remaining infections were detected by either method. Although culture achieved about 10% higher sensitivity, histopathology was the only examination that enabled identification of P. stagnora. These findings align with earlier reports, which highlighted technical challenges in Prototheca diagnostics, particularly when the algal load is low or cell viability is compromised [8,22,60–62]. Accordingly, the combined use of microbiological and histopathological methods improves diagnostic reliability.

Conclusions

To conclude, this study provides new insights into experimentally induced protothecosis in mice, emphasizing how the pathogen species, the inoculation dose and route, and the host immune status all impact the development of infection. The findings from this investigation can be distilled into four key messages. First, pathogenic capacity differs between species, with P. ciferrii being the most virulent and P. stagnora showing the lowest virulence. Second, host immune competence affects the development and spread of Prototheca infection, with athymic mice being more susceptible to infection and to multifocal involvement over wt mice. Third, higher inocula (10⁷) were associated with higher infection yield, whereas the route of administration influenced the site of infection but not its severity. Fourth, the combined use of microbiological and histopathological methods enhanced the detection of Prototheca algae in host tissues. Overall, this study underscores three features of the protothecal disease, namely species-specific pattern of pathogenicity, chronic, asymptomatic or asymptomatic infection as a predominant clinical manifestation, and a pivotal role of the host immunological background in shaping the course and severity of the disease. The findings from this work provide a template for future animal model studies, advancing our understanding of Prototheca pathogenicity and guiding strategies to investigate the virulence mechanisms of this peculiar pathogen. Future research is expected to elucidate the mechanisms of Prototheca virulence and host-pathogen interactions. The use of different immunodeficient models is thought to provide a better understanding of the roles of adaptive and innate immunity in protothecosis.

Supporting information

Acknowledgments

The authors would like to thank Agnieszka Kwiatek, PhD, Mateusz Iskra, PhD, Patryk Mazur, MSc, and Krzysztof Zakrzewski, MSc for their valuable technical assistance in the laboratory. The authors are also grateful to Elżbieta Budzińska-Wrzesień, PhD, for her support with ethical approval procedures and helpful advice on animal welfare.