Impact of Plasminogen on Clostridioides difficile Colitis

Abstract

Pathogens often manipulate the host plasminogen activation system to facilitate escape from the initial site of infection, promote systemic invasion, and evade immune protection. Plasminogen, which is known for its involvement in fibrinolysis, immune modulation, and wound healing, has recently been linked to Clostridioides difficile spore germination. By identifying the mechanisms that drive spore germination and their contribution to C. difficile pathogenesis, in animal model studies, we gain insights into their role in disease severity and mortality. Moreover, inhibiting plasminogen activation using a specific single-chain variable fragment (scFv) has shown the potential to reduce infection severity and mortality. These findings suggest that targeting plasminogen-mediated pathways can be a promising therapeutic approach for managing C. difficile colitis.

1. Introduction

1.1 Plasminogen, activation, and function

Plasminogen, the inactive precursor of plasmin, is present in the bloodstream at a high concentration of 2 μM. It is activated into plasmin only upon binding to a substrate surface and co-localizing with plasminogen activators, ensuring spatial and temporal regulation of plasmin generation. Once free in the bloodstream, plasmin is rapidly neutralized by its specific inhibitor, α2-antiplasmin, and the generic protease inhibitor, housekeeping α2-macroglobulin [1], to prevent excessive fibrin degradation that can result in hemorrhage and tissue damage. As well as that, the activation of plasminogen by tissue plasminogen activator (tPA) and urokinase plasminogen activator (uPA) is stringently controlled by plasminogen activator inhibitors 1 and 2 (PAI-1 and PAI-2). In humans, plasmin plays a vital role as a fibrinolytic enzyme and is essential for cell migration, wound healing, tissue remodeling, and the resolution of inflammation [2].

1.2 Plasminogen in infection

Plasmin(ogen) plays a key role in the crosstalk between fibrinolysis and infections [1, 3]. Many pathogens exploit the plasminogen activator (PA) system to cause significant tissue damage while escaping from the host’s infection containment.

Through host-pathogen co-evolution, some pathogens have developed highly specific and high-affinity plasminogen receptors, allowing them to capture plasminogen effectively. Some pathogens have advanced further by expressing specific plasminogen activators, such as streptokinase by Streptococcus and staphylokinase by Staphylococcus, which directly convert bound plasminogen into active plasmin. Targeting plasmin(ogen), a highly abundant plasma protease known for its ability to break down a wide range of substrates, allows pathogens to sustain robust proteolytic activity on their surface. Because the plasmin binds to the surface of pathogen, it evades inhibition by the host’s α2-antiplasmin.

With the surface-bound plasmin(ogen), pathogens can readily escape the physical confines of extracellular matrices and fibrin clots and spread from the initial site of infection. This interaction also leads to plasmin-mediated degradation of critical immune components, such as complement component 3b (C3b). Such degradation reduces the effectiveness of complement-mediated opsonization, thereby compromising the host’s immune surveillance [4].

There are numerous examples of pathogens exploiting the host’s plasminogen system to enhance their virulence, each utilizing it for a variety of invasive tactics. Plasmodium sporozoites, for instance, bind to plasminogen for navigation through the dermis and liver [5]. Similarly, Leishmania promastigotes exploit surface-bound plasminogen for tissue penetration, immune evasion, and nutrient acquisition [5, 6]. Trypanosoma utilizes plasminogen to breach the blood–brain barrier [7], while Schistosoma relies on it for parasite survival [8]. Furthermore, viruses such as Influenza A exploit plasminogen for its replication, causing severe lung inflammation [9, 10] and promoting brain invasion [11]. This strategic manipulation of plasminogen is crucial for the virulence of these pathogens and the successful establishment of infections.

Plasminogen binding is a pivotal strategy among bacterial species, significantly boosting their invasive capabilities. This strategy is observed across both Gram-positive and Gram-negative species, including highly invasive bacteria such as Streptococcus [12, 13] and Staphylococcus [14, 15]. Some of the bacterial receptors exhibit an exceedingly high affinity for plasminogen, far surpassing that of the host’s receptors—some by more than 100-fold. By leveraging plasminogen, these bacteria enhance their infectious potential, facilitating their entry and dissemination [16, 17, 18].

High-affinity plasminogen-binding proteins and receptors have been identified in a diverse array of bacteria (Table 1). These include OspE/F-related Protein A, C and P (ErpA, C, and P) from Borrelia burgdorferi, which are implicated in Lyme disease [19, 20]; Leptospiral Immunoglobulin-like Conserved Protein (LIC), Major Leptospiral Protein 36 (MLP36), and Lipopolysaccharide-Like Lipoprotein 40 (LipL40) from Leptospira interrogans, associated with leptospirosis [21, 22]. Additionally, Group A Streptococcus pyogenes (GAS), responsible for conditions ranging from streptococcal pharyngitis to necrotizing fasciitis, expresses a high-affinity plasminogen-binding M-like protein (PAM) [23, 24]. Furthermore, the glycolytic enzyme fructose 1,6-bisphosphate aldolase from Mycobacterium tuberculosis, which affects over a quarter of the global population, also demonstrates significant plasminogen binding [25]. Notably, DnaK, a heat-shock protein from the non-pathogenic Bifidobacterium animalis subsp. lactis, exhibits a high affinity for plasminogen, though its functional significance is yet to be determined [26].

Pathogen	Disease	Receptor	Binding affinity KD (nM)	Comments	References
Borrelia burgdorferi	Lyme disease	ErpP, ErpC, and ErpA1	~25	Membrane-anchored lipoproteins with >98% identical sequence	[19, 20]
Leptospira interrogans	Leptospirosis	LICs2	~11	Membrane-anchored putative lipoproteins	[21]
Leptospira interrogans	Leptospirosis	MPL363	3.52 ± 3.95	Membrane-anchored lipoprotein	[21, 22]
Leptospira interrogans	Leptospirosis	LipL404	18.24 ± 2.45	Membrane-anchored lipoprotein	[21, 22]
Streptococcus pyogenes	Invasive GAS5 disease	PAM6	~1	Membrane-anchored M-protein	[23, 24]
Mycoplasma tuberculosis	Tuberculosis	Fructose-1,6-bisphosphate aldolase	6.7 ± 3	Peripheral gluconeogenic enzyme	[25]
Bifidobacterium animalis subsp. lactis	Non-pathogenic	DnaK	~12	Peripheral heat-shock protein	[26]

Table 1.

Examples of bacterial proteins with high-affinity for plasminogen binding.

¹

Erp: OspE/F-related proteins.

²

LICs: Leptospiral Immunoglobulin-like Conserved Proteins.

³

MLP36: Major Leptospiral Protein 36.

⁴

LipL40: Lipopolysaccharide-Like Lipoprotein40.

⁵

GAS: Group A Streptococcus.

⁶

PAM: Plasminogen-binding group A streptococcal M protein.

1.3 Infectious colitis and Clostridioides difficile infection (CDI)

Infectious colitis is a chronic inflammatory condition triggered by pathogen infections and represents a significant global health issue. Existing studies highlight the substantial role of the PA system in colitis. Notably, PAI-1 is highly expressed in colitis and exacerbates mucosal damage [27]. Additionally, levels of fecal PAI-1 in patients directly correlate with disease progression and response to treatments [28].

However, the role of plasminogen in the development and progression of colitis is not well-documented. One study underscores that inhibiting plasmin can alleviate colitis in a murine model. The authors demonstrated that administering a plasmin active site inhibitor, YO-2, halts the development of colitis; they attributed this effect to the inhibition of plasmin-mediated activation of matrix metalloproteinase 9 (MMP9), a crucial mediator of inflammatory responses within the host [29].

Despite these insights, how the PA system impacts the pathogenesis and clinical manifestations of infectious colitis remains elusive. To address this gap, we set out to study the role of plasminogen in colitis, particularly that induced by Clostridioides difficile infection [30]. Using a preclinical animal model, we discovered that plasminogen exacerbates disease pathological symptoms and fatality. In the same study, we also unveiled the mechanistic role of plasminogen during infection, offering deeper insights into its impact on disease progression.

Clostridioides difficile infection (CDI) represents a growing global public health issue. In the United States, it is estimated that approximately 500,000 cases occur each year, with a mortality rate of 6% within a month of diagnosis [31]. Furthermore, CDI recurrence is a significant concern, with up to 35% of patients experiencing a relapse after initial recovery [32]. The economic burden of CDI is considerable, amounting to tens of billions of dollars annually in healthcare costs [33].

Clostridioides difficile, formerly known as Clostridium difficile, is a Gram-positive, spore-forming anaerobe that causes severe gastrointestinal infections. CDI manifests with symptoms ranging from diarrhea to abdominal pain, which can progress to life-threatening pseudomembranous colitis. Notably, 15–20% of colon infections following antibiotic therapy are attributed to CDI [34], frequently initiated by disruption of the normal intestinal flora. This disruption permits the germination of C. difficile spores in the gut, leading to bacterial colonization and toxin production. The C. difficile toxins, specifically Toxin A and B (TcdA and TcdB), and C. difficile transferase, inflict damage on intestinal cells and can provoke extensive colonic inflammation [35, 36].

CDI is transmitted primarily through the fecal-oral route by spores that can withstand extreme conditions such as heat, radiation, and disinfection [37]. These resilient spores persist in environments around the infected individuals, contributing to epidemics in hospitals and residential facilities, community settings, and childcare centers [32, 38]. Figure 1 illustrates the CDI cycle. Ingested spores germinate in the gastrointestinal tract upon exposure to bile salts and in the presence of nutrients such as glycine and alanine. This process involves unpacking the spore’s subtilisin-like proteases, C. difficile proteins A, B, and C (CspA-C), which activate the spore cortex-lytic enzyme (SleC). The degradation of the cortex then triggers the expression of spore germination genes (such as GerG and GerS), leading to the formation of vegetative cells [39]. These vegetative cells colonize at the infection site and cause extensive tissue damage. Spores produced during CDI are responsible for dissemination to other organs, and shedding of spores from patients leads to the spread of CDI. Furthermore, patients who have recovered from C. difficile will face a significant risk (over 35%) of recurrence due to the persistence of these spores within their gastrointestinal tract [32, 40, 41]. Therefore, managing and preventing CDI represent substantial challenges.

Figure 1.

Clostridioides difficile infection cycle. Clostridioides difficile infection begins when spores are ingested. Disruption of the normal gut flora, often due to antibiotic treatment, allows these spores to germinate into active bacteria. The vegetative bacteria then produce toxins that cause inflammation and damage the gut epithelium. Spores formed during the infection can remain dormant inside the host even after recovery and can be shed, potentially initiating new infection cycles.

1.4 Human plasminogen exacerbates C. difficile infection

In our investigation into the effects of plasminogen inhibition on bacterial infections, we established an animal model for CDI using C57BL/6 J mice [30]. We administered M7404 C. difficile spores—a Canadian human epidemic isolate—via oral gavage at 10^6 spores per animal. Human plasminogen was administered intraperitoneally at 3 μM, one hour before and seven hours after infection. We observed deteriorating infection scores, assessed through cage appearance and fecal consistency, along with increased spore counts in the kidney, spleen, and thymus. Crucially, the survival time of infected mice decreased from 48 hours to 24 hours in the presence of human plasminogen [30]. Therefore, our findings indicate that human plasminogen significantly exacerbates the severity of the infection.

To elucidate the mechanism by which human plasminogen promotes CDI, we investigated its binding affinity to C. difficile. Utilizing Western blot analysis with an anti-human plasminogen-specific antibody (MAB2596; R&D Systems, Minneapolis, USA), we found that human plasminogen binds exclusively to the spores but not to the vegetative cells (Figure 2). This specific binding to the spores was consistently observed across other C. difficile strains from various epidemics, such as R20291 [42] and VPI 10463 (Clostridioides difficile (Prevot) Lawson et al. (ATCC 43255)) [30]. BIAcore showed that the binding affinity (K_D) of human plasminogen to C. difficile spores is exceptionally high at 13.4 ± 3.2 nM. Notably, rodent plasminogen did not exhibit detectable binding to the C. difficile spores, underscoring the species-specificity and highlighting the complex co-evolutionary dynamics of this host-pathogen interaction. Confocal imaging further substantiated these findings by showing co-staining of C. difficile spores with fluorescently labeled antibodies specific to human plasminogen (Invitrogen, San Diego, USA).

Figure 2.

Human plasminogen binds to C. difficile spore. (A) Human plasminogen binds to the spores but not the vegetative cells determined by Western using an anti-plasminogen antibody (MAB2596; R&D systems, Minneapolis, USA). (B) the affinity of human plasminogen for the C. difficile spore immobilized on a C1 chip (Cytiva, USA) [30] was determined by BIAcore using single-cycle kinetics in PBS at 20 μL/min with a 60-second contact time and a final dissociation time of 10 min to be 13.4 ± 3.2 nM fitted with a one-site specific binding model. However, mouse plasminogen does not show any binding at the concentration range 0–2.5 μM under the same experimental conditions. (C) Confocal microscopy image shows human plasminogen (red, Invitrogen, San Diego, USA) binding to the C. difficile spore (green, Invitrogen, San Diego, USA) [30]. The spore antibody (green) and plasminogen (red) bind to different epitopes on the spore and the signals do not overlap.

In the study, we further revealed that human plasminogen-bound spores isolated from infected animals are converted into enzymatically active plasmin. Given the lack of a PA within the C. difficile genome, we hypothesize that this activation is mediated by a host PA present at the infection site [30].

Remarkably, treating C. difficile with plasmin alters the spore surface appearance as illustrated by the Transverse Transmission Electron Microscopy images (TEM, Figure 3B–C). Specifically, plasmin treatment leads to a significant reduction in the thickness of the outermost exosporium layer, decreasing from an average of 74 nm to approximately 35 nm—less than half its original size (Figure 3D). Furthermore, we demonstrated that this surface remodeling markedly enhances spore germination, as evidenced in Figure 3E [30].

Figure 3.

Remodeling of the C. difficile exosporium by human plasmin. (A) a cartoon illustration of the C. difficile spore structure. TEM images showing the exosporium of a C. difficile spore without (B) and with (C) human plasmin treatment, where the exosporium is marked with a white double arrow and a black double arrow, respectively, with the average length shown in panel (D). (E) Spore germination (OD₆₀₀(t)/OD₆₀₀(t0)) was determined after 30 min of incubation in the presence of the bile salt taurohyodeoxycholic acid (Tauro) with and without human plasmin-treatment (Plm). Also shown is the spore-alone control [30].

This study provides the first evidence of how human plasminogen binds to C. difficile spores and promotes their germination. Although the binding of human plasminogen to spores of other organisms, such as Bacillus anthracis, has been previously documented, it does not directly influence spore germination in those instances [43]. Current literature lacks any information to further elucidate these findings, particularly regarding whether plasminogen binds to a specific protein within the exosporium or to the composite exosporium surface, which facilitates such high-affinity interactions with a K_D measured in the low nanomolar range. Structurally, the exosporium layer comprises distinct groups of proteins, intercalated to form this complex and resilient structure; among these proteins are the Bacterial Collagen-like Protein A1-A3 (BclA1-A3) and the cysteine-rich morphogenetic factors C difficile Exosporium Protein A-C and M (CdeA-C and CdeM) [44, 45].

It is essential to highlight the disparity in plasminogen concentration: while it’s relatively high in plasma, approximately 2 μM, it remains negligible in the lower gastrointestinal tract. However, this situation changes when the gastrointestinal tract is damaged or infected. Such conditions compromise the permeability of the enteric epithelium, resulting in a substantial increase in plasminogen levels at the injured site. Here, we propose that during infection vegetative cells of C. difficile release toxins, including TcdA, TcdB, and CD transferase, which can cause significant epithelial damage. This damage subsequently leads to plasminogen accumulation at the infection site. Human plasminogen binding to C. difficile spores promotes their germination and spread (see Figure 4). Together, human plasminogen exacerbates the progression of C. difficile infections within this model.

Figure 4.

Working model on the inhibition of human plasminogen activation protects mice from C. difficile infection. C. difficile Infection leads to damage of the gut epithelium and plasminogen accumulation at the infected site. Plasminogen binds to the spores and facilitates spore germination. scFv-A inhibits plasminogen activation and attenuates C. difficile infection [30].

1.5 Inhibition of human plasminogen activation on CDI

We next investigated the effect of inhibiting human plasminogen activation on CDI. For this purpose, we employed a single-chain variable fragment (scFv) named scFv-A, which exhibits a high binding affinity for human plasminogen (K_D = 0.2 nM, Figure 5B) [30]. An X-ray crystallography study provided structural insights into the molecular interactions underlying the inhibitory function, revealing that the CDRH3 region of scFv-A binds to the activation loop of human plasminogen (₅₅₄EPKKCPGR↓_{activation site}VVGGCVAH₅₆₉) as highlighted in bold (Figure 5F). Based on this finding, we proposed that scFv-A inhibits human plasminogen activation by obstructing the binding and activation of PAs (Figure 5A). We confirmed this hypothesis by demonstrating that scFv-A effectively blocks the plasminogen activation mediated by tissue PA (tPA, Figure 5D) with an IC₅₀ value of 42 nM, and urokinase PA (uPA, data not shown). Notably, the inhibition by scFv-A does not affect plasmin’s activity, as shown in Figure 5E.

Figure 5.

Targeted plasmin(ogen) inhibition. (A) Inhibition of plasmin function can be achieved by inhibiting plasminogen activation or plasmin activity. The affinity of human plasminogen for scFvs immobilized on a NiHc chip (Xantec, Germany) was determined by BIAcore using single-cycle kinetics in PBS at 20 μL/min with a 180 second contact time and a final dissociation time of 10 min and fitted with the 1:1 Langmuir model [30] revealed (B) high-affinity binding of scFv-A for human plasminogen (K_D = 0.2 nM, dark green) and a moderate affinity for human plasmin (K_D = 2.5 nM, gray); and (C) high-affinity binding of scFv-I to both human plasminogen and plasmin (K_D = 4.8 and 18 pM, and, deep purple and gray, respectively). (D) Human plasminogen (20 nM) activation by tissue plasminogen activator measured in the presence of 20 mM of epsilon aminocaproic acid (EACA) and scFv-A and scFv-I at 0–200 nM. The IC₅₀ for scFv-A is 42 $\pm$ 7 nM and for scFv-I is 53.1 $\pm$ 8 nM. (E) Human plasmin (20 nM) activity measured in the presence of 0–200 nM of scFv-A or I. While scFv-A does not inhibit the activity of human plasmin at all, the IC₅₀ for scFv-I is 24 $\pm$ 1.4 nM. The fluorogenic plasmin substrate is D-Val-Leu-Lys-AMC, Bachem, at 200 μΜ) and (F) X-ray crystal structure of scFv-A (variable light chain in magenta and heavy chain in deep purple) and human miniplasminogen (consists of Kringle 5, deep teal, activation loop, yellow, and catalytic domains, cyan); top and bottom panels show residues (labeled) involved in the intermolecular interaction between the CDRH3 with the catalytic domain, and the variable heavy chain with Kringle 5 domain, colors as before [30].

To evaluate the efficacy of scFv-A in combating CDI in vivo, it was administered intraperitoneally at a dosage of 40 mg/kg to animals at two critical time points: one hour before and seven hours after infection with C. difficile spores. The study also included three control groups: a buffer control, a naïve scFv of the same scaffold (scFv-naïve), and scFv-I, a high-affinity active site scFv inhibitor (K_D of 0.02 nM). Additionally, a group receiving a combination of both scFv-A and scFv-I was included. The impact of scFv-A on CDI was monitored by physiological appearances; fecal pellets and tissue biopsies were collected for analysis. Viability was assessed in 24- and 28-hours post-infection. The results of this study are presented in Figure 6 [30].

Figure 6.

Inhibition of human plasminogen activation protects mice from C. difficile infection. (A) a schematic shows the experimental layout. The impact of C. difficile infection (10^6 spores orally) with Hepes buffer (black), scFv-naïve (red), scFv-I (green), scFv-A (purple), and scFv-A & scFv-I combined (light purple) is recorded 24 hours post-infection on (B) physiological appearance (3 being most severe), (C) cage appearance (4 being most severe), and fecal consistency (3 being most severe) and (D) survival at 24 and 48 hours [30].

Remarkably, the groups treated with scFv-A, as well as the combined scFv-A and scFv-I groups, exhibited significant protection against CDI. These beneficial effects were evidenced by delayed disease onset, improved fecal consistency, reduced nest soiling, a healthier physiological appearance, and notably, higher survival rates 24 hours post-infection (Figure 6B). Collectively, these findings suggest that scFv-A, which inhibits plasminogen activation, effectively rescues hosts from CDI, likely via preventing the plasmin-driven remodeling of the exosporium, thereby inhibiting spore germination and spread.

2. Conclusion and discussion

Many bacterial pathogens utilize the human plasminogen activation system to breach host tissue barriers. Our research, utilizing a C. difficile colitis animal model, has demonstrated for the first time that human plasminogen binds with high affinity to C. difficile spores. Figure 4 illustrates that during CDI, vegetative cells produce bacterial toxins, including TcdA, TcdB, and CD transferase, causing extensive damage to epithelial cells. This damage and the resulting inflammation led to an accumulation of plasminogen at the infection site, where it adheres to the spores. Once bound, plasminogen is converted to plasmin, which remains attached to the spore, avoiding inhibition by α2-antiplasmin. This process thins the outermost exosporium layer, aiding spore germination and dissemination within the host. The deleterious effects of human plasminogen in facilitating CDI are highlighted by the effectiveness of scFv-A, a high-affinity scFv that inhibits plasminogen activation and significantly improves infection outcomes and survival rates. These findings underscore scFv-A’s potential as a therapeutic agent in preclinical studies (Figure 4).

However, the active site inhibitor, scFv-I, was intriguingly and unexpectedly lacking any efficacy, prompting the need for further investigation. We will explore whether its route of administration results in inadequate concentrations at the infection site or if competition occurs between the spores and scFv-I for binding to plasminogen. The absence of enhanced protection when combining scFv-A and scFv-I confirms that scFv-A, not scFv-I, provides protection against CDI in this model.

The limitations of our animal model, which relies on the mouse plasminogen activation system to mimic human plasminogen functionality, present another challenge. This model does not fully account for the complexities of human plasminogen regulation by activators and inhibitors such as tissue plasminogen activator, urokinase, PAI-1, PAI-2, and α2-antiplasmin. More comprehensive studies using models that incorporate all components of the plasminogen activation system from the same species are crucial to fully understand the impact of plasminogen on CDI.

Additional constraints include the delivery method of scFvs. Intraperitoneal administration may not fully exploit the potential of the antibodies. Techniques such as microencapsulation, nanoparticle carriers, and bio-adhesive polymer formulations could improve targeted delivery directly to the lower gastrointestinal tract. These methods are expected to enhance the efficacy of treatments like the therapeutic C. difficile toxin-neutralizing antibody ABAB-IgG1 [46, 47, 48].

Another limitation is the short half-life of the antibody fragments, ranging from minutes to hours [45]. To address this, we proposed reformatting scFv-A into an Fc-fusion protein [49], which is expected to extend its half-life to approximately one month and substantially enhance its efficacy. Ongoing research focuses on measuring antibody concentrations at the infection site to optimize treatment regimens and evaluate efficacy are warranted.

Finally, identifying the specific receptors involved in plasminogen binding to spores would greatly enhance our understanding of the mechanisms behind spore germination. Such discoveries could provide novel insights into inhibiting human plasminogen during pathogen infections. This knowledge is also crucial for determining whether disrupting pathogen-host protein interactions could be a viable strategy for future therapeutic advancements.

Funding and acknowledgment

James C. Whisstock is supported by an Australian Research Council (ARC) Laureate Fellowship. The authors also wish to acknowledge the support of NHMRC under the Project Grant Scheme APP1127593 and figures were created with BioRender.com.