Review of Immunological Plasma Markers for Longitudinal Analysis of Inflammation and Infection in Rat Models

Abstract

Disease or trauma of orthopedic tissues, including osteomyelitis, osteoporosis, arthritis, and fracture, results in a complex immune response, leading to a change in the concentration and milieu of immunological cells and proteins in the blood. While C-reactive protein levels and white blood cell counts are used to track inflammation and infection clinically, controlled longitudinal studies of disease/injury progression are limited. Thus, the use of clinically-relevant animal models can enable a more in-depth understanding of disease/injury progression and treatment efficacy. Though longitudinal tracking of immunological markers has been performed in rat models of various inflammatory and infectious diseases, currently there is no consensus on which markers are sensitive and reliable for tracking levels of inflammation and/or infection. Here, we discuss the blood markers that are most consistent with other outcome measures of the immune response in the rat, by reviewing their utility for longitudinal tracking of infection and/or inflammation in the following types of models: localized inflammation/arthritis, injury, infection, and injury + infection. While cytokines and acute phase proteins such as haptoglobin, fibrinogen, and α₂-macroglobulin demonstrate utility for tracking immunological response in many inflammation and infection models, there is likely not a singular superior marker for all rat models. Instead, longitudinal characterization of these models may benefit from evaluation of a collection of cytokines and/or acute phase proteins. Identification of immunological plasma markers indicative of the progression of a pathology will allow for the refinement of animal models for understanding, diagnosing, and treating inflammatory and infectious diseases of orthopedic tissues.

1. Introduction

Following injury or infection, an inflammatory response is initiated, as the body attempts to resolve or contain the injurious stimulus. Initially, proinflammatory cytokines, such as interleukin (IL)-1, IL-6, and tumor necrosis factor-α (TNF-α), are produced.^1,2 As these cytokines rise in concentration, they signal production of anti-inflammatory cytokines such as interleukin-1 receptor antagonist (IL-1RA), IL-10, and transforming growth factor-β (TGF-β).¹ The immune response from infection starts with activation of phagocytes and toll-like receptors that then trigger the inflammation pathway.³ Acute or chronic inflammation can occur from infection or injury, and the duration greatly depends on how the body enters the inflammatory phase. Prolonged or overproduction of inflammatory cytokines can lead to tissue damage and/or chronic inflammation. Chronic inflammation is characterized by tissue destruction and/or sustained levels of acute phase proteins (APPs).^3–5 In contrast, acute inflammation begins to resolve after 24–48 hours with rising anti-inflammatory protein concentrations to halt production of APPs.^2,3

Here, we review rat models of localized inflammation/arthritis, injury, infection, and injury + infection, with a focus on musculoskeletal tissues. Many inflammatory and infectious conditions have been successfully modeled in rats^6,7 and can be monitored through inflammation mediators. Rats are often chosen due to their ease of handling compared to larger species and their larger blood volume than mice.⁸ This allows for relatively frequent blood draws that can be analyzed in a longitudinal study without the sacrifice of animals at each time point; however, available blood volume is still very low, limiting the number of markers that can be tested. Maximum recommended blood collection in rats is 7.5% of circulating blood volume in a 7-day period.⁹ To detect statistically significant changes following an inflammatory stimulus, the markers must be sensitive and have little variability. Though established rat models of many injuries and diseases exist, longitudinal studies using these models have reported inconsistent changes in the levels of immunological mediators across similar disease/injury models.^10–12 Identification of sensitive and reliable longitudinal markers will enable quantitative evaluation of the progression and severity of infection or inflammation, for further characterization of these models. Additionally, longitudinal quantification of inflammation is important in determining the efficacy of treatments and offers a more in-depth understanding of the healing process.

Certain plasma markers, such as C-reactive protein (CRP)¹³ and white blood cell (WBC) counts¹⁴ are well accepted as clinical indicators for different conditions of inflammation and infection but are not reliable in certain orthopedic cases.^14,15 Many clinicians will use multiple plasma markers, such as the Laboratory Risk Indicator for Necrotizing Fasciitis (LRINEC) score, which includes blood CRP, WBC, hemoglobin, sodium, creatine, and glucose levels, to help determine the need for a revision surgery; however, this score does not accurately differentiate necrotizing fasciitis and cellulitis.¹⁶ Similarly, recent clinical research has explored biomarkers indicative of pathogen-specific response to infection to improve diagnosis and prognosis using proteomic mapping, specifically for Staphylococcus aureus (S. aureus) bacteremia^17,18 and osteomyelitis.^19–21 The same markers used for diagnosis of human diseases clinically are not necessarily useful in rat models due to differing biological activity and varying concentration changes during immunological response. In the 1980s and 1990s, rat CRP was the standard inflammation marker,^12,22 but studies have since identified markers more indicative of inflammation such as haptoglobin or α2-macroglobulin.^23,24 Interleukins, APPs, and other blood factors have also been evaluated as indicators of infection/inflammation progression, but the same markers are rarely tested across studies making it difficult to determine a standard. Without this standard, longitudinal tracking of inflammation generates limited comparisons between studies and generally requires increased research time and resources.

To determine consensus regarding the efficacy of immunological markers for tracking immunological response to infection and inflammation in rats, manuscripts featuring localized inflammation/arthritis, injury, infection, or injury + infection in rat models were reviewed (Figure 1). To be included, the study’s methodology must have involved either an initial blood collection prior to induction of inflammation/infection or a healthy control. Generally, studies featured either longitudinal blood collections or individual groups for each time point. Blood was analyzed for the presence of various immunological proteins through biochemical assays. The scope was not limited strictly to longitudinal studies because analysis of multiple blood proteins can require a blood volume higher than that which is available. While some articles featured application of a treatment, discussion of these models is limited. Due to the relatively limited exploration of inflammatory markers in rat models of musculoskeletal diseases and injuries, rat models of various inflammatory states were included, to provide a more comprehensive report of biomarker utility in the rat species. Because most rat models are relatively short term (days to weeks), our assessment focused on biomarker utility in models of acute inflammation; chronic inflammation was considered only in models with extended timelines (> 2 weeks). While rat age, sex, and strain were rarely evaluated as factors in the variety of models reviewed here, it is important to note that each of these factors would contribute to differences in the immunological response.²⁵ With this review, our objective was to identify the blood markers that are the most consistent with the immune response to infection and inflammation in the rat. The biomarkers discussed in this article have been previously studied in rat models and present potential relevancy for rat studies. Markers were assessed based on their utility to effectively differentiate longitudinal changes in infection or inflammation, how well they correspond with other outcome measures, and the extent to which they have been tested across rat studies.

Figure 1:

Overview of models reviewed, divided into four categories: models of localized inflammation/arthritis, models of injury, models of infection, and models of injury + infection. ELISAs: enzyme-linked immunosorbent assays. (Created with BioRender.com)

2. Models of localized inflammation/arthritis

Models of localized inflammation or arthritis generally include the injection of an inflammation-inducing chemical, commonly in the hind limb. Turpentine oil, casein, and croton oil are agents that cause localized inflammation. Freund’s complete and incomplete adjuvant are additional agents that can be used to simulate rheumatoid arthritis, an autoimmune disorder in which the immune system attacks the cartilage of the joints.^26,27 Models using carbon tetrachloride were not considered, as it has been shown to induce cirrhosis and an atypical acute phase response.²⁸

2.1. Haptoglobin

Many studies have reported haptoglobin as a sensitive marker of localized or arthritic inflammation in rat models. In Abbak et al., female 3-month-old Wistar albino rats receiving subcutaneous injections of turpentine oil demonstrated increased haptoglobin levels on days 1 and 7 (compared to day 0).¹¹ In female Wistar Hanover rats (176.5 ± 21.5 g), haptoglobin peaked 36 hours (7x control) after intraperitoneal injection with Freund’s complete adjuvant and remained elevated throughout the 21-day study.²⁴ In arthritic models, haptoglobin has been found to reflect disease progression via two distinct peaks: initial, localized inflammation and delayed, systemic inflammatory effects in Sprague Dawley rats (165–190 g) injected with desiccated Mycobacterium butyricum in mineral oil, polysorbate-80, and sodium chloride;²⁹ male Lewis rats injected with Freund’s complete adjuvant;²⁷ and female AHH/R rats injected with heat-killed Mycobacterium tuberculosis in paraffin.³⁰ In an arthritic model using injections of type II collagen in incomplete Freund’s adjuvant in male Wistar rats, haptoglobin levels were increased at days 7, 14, 21, and 28 compared to day 0.⁶ At 6 weeks post-injection, post-mortem haptoglobin levels correlated with terminal Arthritic Index scores of swelling and redness in the hind paws. After daily dosing (excluding weekends) from weeks 3 to 6 (post-injection) with nonsteroidal anti-inflammatory drugs, both haptoglobin levels and arthritic inflammation were reduced. Steroids and immunosuppressives reduced arthritic inflammation but increased haptoglobin levels, while antirheumatics increased arthritic inflammation but had little effect on haptoglobin levels.⁶

2.2. α₂-Macroglobulin

α₂-macroglobulin (α2M) is a commonly tested APP that has been found to indicate the presence of localized or arthritic inflammation in rat models. Okubo et al. found that α2M reached peak blood levels 1–3 days (no statistical analysis) after injection of turpentine oil (110 to 140-fold over control value) in male and female Wistar-King rats.³¹ Others similarly noted α2M peaked at 48 hours post-injection in male and female 6-week-old CD rats (no statistical analysis),³² male 9-week-old Sprague Dawley rats,³³ and 9-week-old Sprague Dawley rats.³⁴ However, dosing studies using turpentine oil suggest an upper limit of α2M production, as the differences in α2M levels were not proportional to dose of inflammatory agent.^33,34 This suggests α2M is an effective indicator of the presence of inflammation but may not accurately reflect the level of inflammation. There has been limited exploration of α2M in response to chronic inflammation.

2.3. Fibrinogen

Fibrinogen is another APP that can be found in the blood and has shown potential as a marker for both acute and chronic inflammation.^{10,11,24,27,35,36} In a chronic arthritic model using female 10-week-old Dark Agouti rats, fibrinogen peaked at day 14 following injection of collagen in Freund’s incomplete adjuvant and did not return to baseline until day 70, making it both an early marker of joint inflammation and an indicator of chronic inflammation.³⁵ Myers and Fleck noted a 9-hour delay between injection of turpentine and detectable increase of fibrinogen in the blood, with fibrinogen levels reaching peak values between 18 and 24 hours (no statistical analysis) in male Sprague Dawley rats (305–360 g).¹⁰ Additionally, Abbak et al. found that female 3-month-old Wistar albino rats injected with turpentine oil displayed increased levels of fibrinogen through the 7-day study.¹¹ In a model using Freund’s complete adjuvant, fibrinogen levels were increased at 12 hours post-injection and peaked at 24 hours in female Wistar Hanover rats (176.5 ± 21.5 g).²⁴ While levels returned to normal on day 4, a second peak occurred on day 7.²⁴ Similarly, fibrinogen levels peaked at day 5 and then again between days 7 and 15 following injection of Freund’s complete adjuvant in male Lewis rats.²⁷

2.4. C-Reactive Protein

There is much debate on the functionality of CRP as an indicator of localized or arthritic inflammation in rats. Of note, baseline levels of CRP in rats (300–600 μg/mL)^10,22 are higher than those in humans (~1000 μg/L).¹² Injections of male and female rats with casein or croton oil caused only a 2- to 3-fold increase in CRP levels at 24 hours post-injection.²² In 1987, Bürger et al. concluded that because CRP increased 1–3 days after turpentine injection and in two arthritic models (induced by Freund’s complete adjuvant injection or 6-sulphanilamidoindazole orally) in 12–14-week-old rats, it was a useful marker to study inflammation.¹² Male Sprague Dawley (280–320g) rats injected with turpentine oil displayed relative increases to 2- to 3-fold initial CRP levels at 48 hours post-injection (no statistical analysis).¹⁰ However, more recent studies using 3-month-old female Wistar albino rats injected with turpentine oil did not detect changes in CRP during the 7-day study; rather, markers such as haptoglobin and fibrinogen were more sensitive.¹¹

2.5. Cytokines

IL-6 is one of the cytokines that regulates APP synthesis and has been used to estimate the extent of localized or arthritic inflammation in rats. IL-6 has been found to peak 12 hours post-injection of turpentine oil in 6-week-old male and female CD rats (no statistical analysis)³² and in 9-week-old Sprague Dawley rats.^33,34 In a 40-hour study, injection of Freund’s complete adjuvant in female Wistar Hanover rats (136.0 ± 10.0 g) resulted in increased IL-6 at 5, 10, 15, 20, 25, and 35 hours compared to pre-injection values.²⁴

IL-1β is another pro-inflammatory cytokine that triggers the acute phase response. IL-1β peaked at 28 times the control level after 24 hours post-injection with Freund’s complete adjuvant in female Wistar Hanover (176.5 ± 21.5 g) rats.²⁴ In contrast, Jinbo et al. reported no changes in IL-1β levels (no statistical analysis) in response to inflammation caused by turpentine oil in 6-week-old male and female CD rats.³² Additionally, IL-2, IL-4, IL-8, and IL-10 were tested in this study. Of these, only IL-8 displayed relative changes, peaking at 12 hours post-injection of turpentine oil, demonstrating its potential for further exploration as a marker of the early inflammatory response.³² IL-2, IL-4, and IL-10 have not shown utility in tracking immunological response to localized inflammation.³²

Additional cytokines explored in models of localized or arthritic inflammation in rats include interferon-γ (IFN-γ),³² TNF-α,²⁴ and cytokine-induced neutrophil chemoattractant-1 (CINC-1).^33,34 In a 40-hour study, female Wistar Hanover rats that received injection of Freund’s complete adjuvant displayed a 2-fold increase in TNF-α at 10 hours post-injection, after which it returned to baseline (15 to 40 hours).²⁴ In a follow-up study carried out 21 days, TNF-α levels at all time points were equivalent to basal levels.²⁴ Based on these studies, TNF-α shows limited utility as an early marker of inflammation and may not be suitable for longitudinal inflammation tracking in long term studies. In 9-week-old male Sprague Dawley rats receiving multiple injections of turpentine oil (0, 24, and 72 hours), differences in CINC-1 based on dose (0.05 or 0.4 mL/rat) were seen at 12, 24, 48, and 96 hours following the first injection but were not consistent throughout the study.³³ For both doses, CINC-1 peaked at 12 hours following the first injection.³³ In a similar study using singular turpentine oil injections (0.05, 0.2, or 0.4 mL/rat) in 9-week-old Sprague Dawley rats, CINC-1 levels peaked at 12 hours post-injection, and while there were significant differences between the lower dose and the two higher doses, there were no significant differences between the two higher doses at any time point.³⁴

Levels of IFN-γ in 6-week-old male and female CD rats did not change in response to injection of turpentine oil (no statistical analysis).³² Though it is a regulator of the acute phase response,³⁷ changes in IFN-γ levels may not be detectable during localized inflammation.

2.6. Serum Amyloid A and Serum Amyloid P

Serum amyloid A (SAA) and serum amyloid P (SAP) are additional markers that have been explored in localized inflammation models in rats because of their demonstrated utility in mice.^10,38 SAP in rats differs in structure from the human counterpart but has similar baseline levels in both species.^22,39 Surprisingly, SAP levels were unchanged throughout the 6-day study in male and female rats injected with croton oil or casein (no statistical analysis).²² Likewise, SAA levels in 3-month-old female Wistar rats receiving turpentine oil showed no changes over the 7-day study,¹¹ indicating that SAP and SAA may not be sensitive enough to differentiate acute inflammation in rats.

2.7. Additional Markers Tested

A variety of additional markers have been tested in models of localized or arthritic inflammation in rats with varying degrees of success, including albumin,^11,24,30 ceruloplasmin,^11,30 α1-acid glycoprotein (AAG),^29,33,34 seromucoid,^29,30 and lipocalin-2 (LCN-2).³⁸

In 3-month-old female Wistar albino rats receiving turpentine injections, albumin levels decreased on days 4 and 7 compared to days 0 and 1.¹¹ Ceruloplasmin was increased on day 4 in rats receiving turpentine or saline, and remained elevated on day 7 in the turpentine group while the saline group returned to baseline.¹¹ In a 40-hour study, female Wistar Hanover rats that received injection of Freund’s complete adjuvant displayed decreased albumin levels at 10 hours, which returned to baseline at 15 hours and decreased again through 40 hours (excluding the 35-hour mark).²⁴ In a follow-up study carried out 21 days, Albumin levels decreased at 12 hours, returned to baseline at 24 and 36 hours, and decreased again from 48 hours to day 14 before returning to baseline at day 21.²⁴ While ceruloplasmin and albumin levels changed in response to increased inflammation, haptoglobin and fibrinogen were found to be more sensitive and reflective of inflammatory state.

In 9-week-old male Sprague Dawley rats receiving multiple injections of turpentine oil (0, 24, and 72 hours), levels of AAG peaked 48 hours after the first injection (59-fold pre-injection).³³ Differences in AAG based on dose of turpentine oil (0.05 or 0.4 mL/rat) were seen at only some time points (24, 48, 120, and 144 hours), and the differences were not proportional to dose.³³ In a similar study using singular turpentine oil injections (0.05, 0.2, or 0.4 mL/rat) in 9-week-old Sprague Dawley rats, AAG levels peaked at 48 hours in all groups.³⁴ AAG levels for the lowest dose were lower than AAG levels for the middle or highest dose at all time points post-injection.³⁴ However, there were no differences in AAG levels between the 0.2 and 0.4 mL/rat doses.³⁴ Levels of AAG and seromucoid displayed similar trends to haptoglobin in an arthritic model using adjuvant in Sprague Dawley (165–190 g) rats, with an initial peak at 2 days and another peak at 14 days (no statistical analysis), corresponding to the known phases of arthritic inflammation.²⁹ While AAG displayed increases in response to increased inflammation, it appears the marker is not sensitive enough to distinguish between dose-dependent inflammatory responses.

LCN-2 blood levels in 8-week-old male Wistar rats increased starting at 12 hours post-injection with turpentine oil and were highest at 48 hours, but the study was not continued further.³⁸ Additional exploration is needed to determine the full response time and sensitivity of LCN-2.

3. Models of injury

Injury induces an immunological response from acute tissue damage incurred during surgical procedures such as laparotomy,³¹ castration,²³ or implantation of a biomaterial. Additionally, inflammation can be induced via bone fracture⁴⁰ or creation of segmental bone defects².

3.1. α₂-Macroglobulin

α2M has demonstrated utility as a marker of surgically-induced inflammation. Male Wistar King (200–250 g) rats subjected to laparotomy displayed relatively increased levels of α2M, which peaked 24 hours post-surgery at approximately 100 times pre-surgery levels (no statistical analysis).³¹ α2M levels peaked at day 2 following castration (4–25 times pre-surgery levels) or oophorohysterectomy (7–28 times pre-surgery levels) in male and female CD rats, respectively (no statistical analysis).²³

3.2. C-Reactive Protein

CRP has been studied in models of various abdominal surgeries (hepatectomy, splenectomy, nephrectomy, colectomy, and gastrectomy) in female Wistar albino rats.⁴¹ Of these procedures, rats that underwent a colectomy exhibited CRP levels elevated over the control, gastrectomy, hepatectomy, and nephrectomy at 48 hours post-surgery, possibly due to the release of bacteria during colon resection.⁴¹ Because of limited study of CRP, other markers such as α2M and TNF-α are recommended over CRP for longitudinal tracking of physically-induced inflammation.

3.3. Cytokines

IL-1β and IL-10 have been explored in response to tibial fractures.⁴⁰ Fractures were created in Sprague Dawley rats using an adjustable drop tower-blade apparatus using a 3-mm, 10-mm, or 15-mm buffer disc, which controlled the depth the blade penetrated the limb.⁴⁰ For all fracture groups, a 500 g blade was dropped from 50 cm. Increasing buffer disc size was associated with more severe fractures and vascular injuries.⁴⁰ The simple fractures (3-mm group) were stabilized with a cast, while the complex fractures (10-mm and 15-mm groups) were stabilized internally with Kirschner wires (K-wires).⁴⁰ All fracture groups displayed increased levels of pro-inflammatory IL-1β compared to uninjured control at days 1, 3, and 7.⁴⁰ Surprisingly, anti-inflammatory IL-10 was increased in all surgical groups at day 1 and remained evaluated over the control throughout the 28-day study.⁴⁰

Additionally, IL-6 has been studied in response to various abdominal surgeries (hepatectomy, splenectomy, nephrectomy, colectomy, and gastrectomy) in female Wistar albino rats.⁴¹ When averaging IL-6 levels across all surgery procedures, levels were highest at 6 hours post-surgery (compared to 24 and 48 hours).⁴¹ The factors surgery and time displayed overall effects, and there was an interaction between the two factors.⁴¹ At 6 hours, rats with gastrectomy had the highest IL-6 levels (elevated over the splenectomy, colectomy, laparotomy (sham), and uninjured control), but at 24 hours, rats with hepatectomy had the highest IL-6 levels (elevated over the splenectomy, colectomy, sham, and control).⁴¹ At 48 hours, only gastrectomy and hepatectomy had levels higher than the sham group.⁴¹ Further exploration of IL-6 in response to surgically-induced inflammation may help to determine its sensitivity to differentiating inflammatory stimuli.

Additionally, TNF-α and TGF-β were studied in response to tibial fractures in Sprague Dawley rats caused by the force of dropping a 500 g blade from 50 cm.⁴⁰ The extent of fracture was limited by various sized buffer discs (3-mm, 10-mm, and 15-mm).⁴⁰ All fracture groups had higher TNF-α levels than the uninjured control group on days 1 and 3.⁴⁰ Further, the most severe fractures (associated with the 15-mm disc) resulted in TNF-α levels higher than those of the other fracture groups on days 7 and 14.⁴⁰ Decreases in TNF-α levels in the fracture groups began on day 14, and the most severe fractures (15-mm group) were associated with the longest elevated TNF-α levels, which remained higher than all groups at day 14, but were equivalent to all groups by day 28.⁴⁰ TGF-β levels remained unchanged for the first week post-surgery, but TGF-β in all fracture groups was higher than in the uninjured control on days 14 and 28.⁴⁰ The results of this study suggest the magnitude and duration of TNF-α are reflective of the inflammatory response due to injury. TGF-β displayed a delayed response to inflammation, which may limit its utility in shorter studies. TGF-β1 was also studied in 8-week-old male Sprague Dawley rats following tibial fractures (transverse fractures created using orthopedic wire saw and stabilized with K-wire) with and without brain injury (resulting from the free fall of a 20 g weight onto the exposed dural membrane underneath the parietal bone).⁴² In the fracture group, TGF-β1 levels were elevated over the uninjured control group at 3 days and 1, 2, and 4 weeks.⁴² Additionally, in the combination fracture and brain injury group, TGF-β1 was elevated over the fracture and control groups at 3 days and 1, 2, and 4 weeks.⁴²

3.4. Additional Markers Tested

Additional markers tested in injury models include cortisol⁴¹ and WBC counts.⁴¹ Female Wistar albino (260–280 g) rats receiving different abdominal surgeries (hepatectomy, splenectomy, nephrectomy, colectomy, and gastrectomy) showed a decrease in WBCs at 24 hours post-surgery compared to 6 hours in all groups except for splenectomy, which did not demonstrate this decrease.⁴¹ In the same model, cortisol showed little to no change following surgical procedures, regardless of elapsed time or procedure.⁴¹ In surgically-induced inflammation models, cortisol does not appear to be a sensitive indicator of inflammatory response. Additionally, WBC levels appear to change in response to surgical procedures but are not sensitive enough to distinguish the extent of inflammation.

An additional model involved testing a panel of cells, cytokines, and chemokines to understand the immune response during bone healing in 13-week-old Sprague Dawley rats with critically sized (8 mm) femoral defects.² Through multivariate analysis of responses in rats receiving either immediate or delayed (at 8 weeks) treatment with bone morphogenetic protein-2 (BMP-2) in alginate hydrogel and a polycaprolactone (PCL) mesh, Cheng et al. found B cells, IL-6, IL-13, and IL-1α correlated with successful healing.² Additionally, IFN-γ induced protein-10 (IP-10), myeloid-derived suppressor cells (MDSCs), and cytotoxic T cells were indicative of poor healing. At week 9 (1 week after delayed treatment), MDSCs, monocytes, IP-10, and IL-1β correlated with poor healing, and B cells, T helper cells, and IL-13 correlated with successful healing. Overall, rats with poor healing displayed a dysregulated immune response.²

4. Models of infection

Models of infection cause an immunological response through inoculation with a pathogen. In animal models, S. aureus is a commonly used pathogen due to its clinical relevance in osteomyelitis, dermatitis, and endocarditis.⁵ While the rats models that fit the inclusion criteria for this review each used S. aureus, it is important to note that the immunological response is pathogen specific and will differ between pathogens.⁴³

4.1. Haptoglobin

Haptoglobin levels have been reported elevated in infected rats. Three-month-old female Wistar albino rats experiencing infection due to subcutaneous injection of S. aureus displayed increased levels of haptoglobin on days 1 and 7 when compared to day 0.¹¹ Notably, day 1 levels of haptoglobin were higher in infected rats compared to rats with inflammation from injection with turpentine oil.¹¹ This discrepancy could correspond with differences in the speed of the acute phase response to infection versus sterile inflammation.⁴⁴

4.2. α₂-Macroglobulin

α2M levels in male and female CD rats inoculated with S. aureus intradermally were relatively increased by day 1 post-injection and peaked at day 2, at 8–33 times the pre-infection levels (no statistical analysis).²³ Similarly to the effects in haptoglobin as a function of the method by which inflammation was induced, peak levels of α2M were lower in this infection model compared to peak levels in localized inflammation/arthritis models of inflammation from turpentine oil.^23,31,45 Again, this difference may be due to differences in the inflammatory stimulation from infection and inflammation. Nonetheless, the results of this study indicate the potential utility of α2M as an early marker of infection.

4.3. Fibrinogen

Fibrinogen levels were found to increase in 3-month-old female Wistar albino rats injected with S. aureus subcutaneously, with levels peaking on day 1, at 3-times the day 0 value.¹¹ Fibrinogen levels were higher than the control (saline) at all study time points (days 1, 4, and 7),¹¹ suggesting fibrinogen may be a sensitive marker of (at least) acute infection.

4.4. C-Reactive Protein

Three-month-old female Wistar albino rats infected from subcutaneous injection of S. aureus displayed no changes in CRP levels, with large intra-group variability.¹¹ While CRP levels reflect levels of infection, obtaining statistically significant results is limited by the small effect sizes and high variability; therefore, other markers, such as haptoglobin and fibrinogen, were determined to be more sensitive than CRP in this model of infection.¹¹

4.5. Serum Amyloid A

Like CRP, SAA displayed no change in response to infection from subcutaneous injection of S. aureus in 3-month-old female Wistar albino rats.¹¹ Other makers in this study including haptoglobin and fibrinogen were shown to be more sensitive to infection.¹¹

4.6. Additional Markers Tested

Ceruloplasmin and albumin were also tested in 3-month-old female Wistar albino rats injected with S. aureus subcutaneously.¹¹ Ceruloplasmin increased at days 4 and 7 in rats receiving S. aureus and only at day 4 in rats receiving saline (control), compared to pre-injection values (day 0).¹¹ When comparing method of inducing inflammation, no differences in ceruloplasmin levels between infected rats and rats receiving turpentine oil were observed on day 7. Albumin levels in infected rats showed no change from pre-infection values.¹¹

5. Models of injury + infection

Models with a contaminated implant or other surgically-induced infection present special cases where there is an immunological response due to both the surgery/implantation and the subsequent infection, such as osteomyelitis, where bone becomes infected and inflamed.^4,5 The majority of human osteomyelitis cases are associated with the presence of Staphylococcal strains.^46,47 Models of injury + infection present a significant clinical relevance to infections secondary to implants, surgery, and fractures; however, exploration of immunological markers in rat models is limited to date.

5.1. Haptoglobin

Haptoglobin has been used to track infection and success of treatment in 5-month-old male Sprague Dawley rats receiving femoral injections of methicillin-susceptible S. aureus (MSSA) into the intermedullary cavity.⁴⁸ Rats receiving a gentamicin-coated K-wire implant with intraperitoneal (IP) gentamicin sulfate injection (combination group) showed decreased levels of haptoglobin on days 7, 28, and 42 compared to uncoated implant alone (control).^48,49 From cultures of the explanted K-wire at days 28 and 42, both treatments (IP injection alone and combination group) displayed reduction of bacteria compared to the control, but there was no difference between treatment groups.^48,49 In this case, haptoglobin levels appeared indicative of (ultimate) bacterial load, suggesting that haptoglobin levels are somewhat reflective of extent of infection. Haptoglobin was also studied in 13-week-old female CD rats using an established implant-based osteomyelitis model⁵⁰, which involves the drilling of a bicortical defect in the mid-diaphysis of the femur and the implantation of a screw soaked in S. aureus (ATCC 6538-GFP). At day 7, the infection site was debrided, and rats were treated with blank chitosan hydrogel, fosfomycin-loaded (3 mg) chitosan hydrogel, or fosfomycin-loaded (3 mg) chitosan hydrogel surrounded by a polycaprolactone scaffold (n=4). Levels of haptoglobin in the chitosan hydrogel and fosfomycin-loaded chitosan hydrogel with scaffold groups were increased by day 3 (post-infection) and remained elevated at day 6 (Figure 3). A further spike in haptoglobin was observed at day 8 (day after treatment surgery), and all groups were elevated over baseline levels at day 10. All groups had returned to baseline by day 21.

Figure 3:

Haptoglobin concentration (mg/mL) in rats following infection with S. aureus-soaked screws at day 0, followed by treatment at day 7 with blank chitosan hydrogel (GEL), fosfomycin-loaded chitosan hydrogel (GEL + FOS), or fosfomycin-loaded chitosan hydrogel surrounded by a polycaprolactone scaffold (GEL + FOS + SCAF) (n=4). A one-way repeated measures ANOVA with Tukey’s post-hoc pairwise comparisons was used, with α=0.05. * difference between treatment groups within a time point. The following differences are within a treatment group over time: # different than day −1, & different than day 1, $ different than day 6, % different than day 8, and ^ different than day 10. p<0.05 for all differences reported.

5.2. α₂-Macroglobulin

α2M has demonstrated utility in a model that elicits immune responses from both surgery and infection. Adult male Sprague Dawley rats received a polyethylene plug in the femoral condyle, a titanium screw in the proximal tibia, and an injection of MSSA or saline into the tibial intercondylar canal and joint space. In both the infected and saline groups, α2M increased following surgery but returned to baseline earlier in the saline group (day 14) compared to the infected group (day 28) (Figure 2A).⁵¹ Nonetheless, infected rats continued to display signs of localized infection through other measures such as radiography (Figure 2B), bone loss, histology, and bacterial counts.⁵¹ α2M was significantly increased in the infected group at days 1, 3, 7, 14, and 21 compared to the saline group, indicating α2M has utility as a marker of acute inflammation and infection and may be useful in differentiating immunological stimulus. However, α2M may not be reflective of long-term immunological effects of persistent infection, such as osteolysis.

Figure 2:

A. Concentration of Alpha 2 Macroglobulin (ng/mL) over a 4-week period following knee-replacements in rats infected with 10⁸ CFU S. aureus compared to those receiving saline. Asterisks indicate a significant difference between groups using the Wilcoxon Rank-Sum test (p<0.05). B. Radiographs of rats receiving saline or 10⁸ CFU S. aureus over a 4-week period. Red arrows indicate the presence of septic loosening, and arrowheads indicate peri-implant radiolucency. Adapted from Fan et al.⁵¹

5.3. C-Reactive Protein

In an osteomyelitis model in male Sprague Dawley rats, levels of CRP were higher in infected rats following soft tissue dissection and injection of S. aureus into the bone marrow cavity, compared to rats receiving the sham procedure (no S. aureus).⁵² Additionally, rats treated with induced membrane technique (vancomycin-loaded polymethyl methacrylate bone cement) had lower levels of CRP at days 1, 3, 5, and 7 post-treatment, compared to the untreated infection group.⁵² In a 28-day study of 26-week-old male Sprague Dawley (350–450 g) rats receiving ultra-highly cross-linked polyethylene tibial implants and S. aureus (ORI16_C02N)-coated or sterile titanium femoral implants, CRP levels were increased in infected rats at day 5.⁵³ In these models, CRP levels displayed longitudinal changes as a function of infection and/or treatment; however, CRP displayed inconsistent results in models of localized inflammation/arthritis, models of injury, and models of infection.

5.4. Cytokines

TNF-α levels in male Sprague Dawley rats with S. aureus induced osteomyelitis were higher than rats that underwent a sham procedure.⁵² Similar to the effects of treatment on CRP, rats receiving the induced membrane treatment displayed lower TNF-α levels than the untreated infected group at days 1, 3, 5, and 7.⁵²

Additionally, levels of TNF-α and IFN-γ in 26-week-old male Sprague Dawley rats (350–450 g) were unchanged following implantation of ultra-highly cross-linked polyethylene tibial implants and S. aureus (ORI16_C02N)-coated titanium femoral implants for a total knee arthroplasty.⁵³ IL-10 levels were increased in both the infected and control groups at day 5 and returned to baseline by day 28.⁵³ No differences in IL-12p70 or IL-4 between groups or time points were observed.⁵³ IL-6 was increased in infected rats on day 28 compared to baseline, and IL-10 was increased in both the infected and control groups at day 5 compared to baseline.⁵³ In Sprague Dawley rats (180–220 g) with S. aureus-induced tibial osteomyelitis, IL-2 was increased over the control group at 1, 2, and 4 weeks post-infection.⁵⁴

5.5. Additional Markers Tested

Additional markers tested in models of injury and infection include hemoglobin,^53,55,56 and WBC counts.^52,53,55,56 While two of the models involved implant-based infection, neither saw any change in levels of hemoglobin.^55,56 Therefore, hemoglobin may not be a reliable indicator of infection resulting from a surgical procedure. Lucke et al. found increased levels of WBCs in 5-month-old female Sprague Dawley rats on days 3 and 7 post-infection compared to uninfected rats, but no differences were observed beyond 7 days.⁵⁵ In male Sprague Dawley rats with S. aureus-induced osteomyelitis, WBC levels were higher in infected rats than in those receiving the sham procedure. Additionally, WBC levels were lower in rats receiving the induced membrane treatment than in those left untreated at days 1, 3, 5, and 7.⁵²

In 26-week-old Sprague Dawley rats (350–450g) with ultra-highly cross-linked polyethylene tibial implants and S. aureus (ORI16_C02N)-coated titanium femoral implants for a total knee arthroplasty, WBC and hemoglobin displayed no significant changes over the 4-week study.⁵³

Vantucci et al. measured a panel of cytokines, chemokines, and cells in response to biomaterial infection following the creation of a 2.5 mm femoral segmental defect by soaking a collagen sponge implant in S. aureus.⁵⁷ In this study using 21-week-old female Sprague Dawley rats, epidermal growth factor, macrophage chemoattractant protein-1 (MCP-1), and LIX were upregulated in the rats receiving a sterile collagen sponge (control) at day 1, and IL-2 was upregulated in the infection group at day 1. On day 3, circulating levels of RANTES were increased in the control group over the infection group, whereas IL-10 was elevated in the infection group compared to the control group. By day 7, levels of eotaxin, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-12p70, MCP-1, TNF-α, and RANTES were elevated in the control group compared to the infected group. Similarly at day 14, eotaxin, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-10, IL-12p70, IL-17A, LIX, MCP-1, IP-10, TNF-α, macrophage inflammatory protein-1 Alpha, and RANTES were elevated in the control over the infection group. At day 28, there were no differences between the control and infection groups. Finally, at day 56, levels of IL-4, IL-12p70, IL-17A, and TNF-α were elevated in the infected group over the control. Through multivariate analysis, immune response in non-infected rats correlated with CD3+ T cells, and infected rats were associated with MDSCs, B cells, and IL-10.⁵⁷ It was concluded that the immune response of the control represented a coordinated response with inflammation resolving between days 14 and 28, while the immunological response of the infected group demonstrated immune dysregulation.⁵⁷

6. Discussion and Conclusion

A multitude of cytokines, acute phase proteins, and other markers have been tested in rat models of infection and inflammation (Table 1). Of these, haptoglobin, fibrinogen, and α2M have each demonstrated utility for tracking infection/inflammation progression in rats. However, markers that have been previously used for other species, such as CRP in humans,¹³ and SAA³⁸ and SAP¹⁰ in mice, have shown limited success in rats. While a few markers exhibited potential utility in multiple model types, there is likely not a singular superior marker for use in rats. Additionally, marker sensitivity may depend on model type, as the immune response differs based on inflammatory stimulus. More recently, the analysis of panels of cells, cytokines, and chemokines has demonstrated the potential to generate a more comprehensive understanding of the immunological response to chemical/physical injury and/or infection.^2,57,58

Table 1:

Biological role and scoring of immunological plasma markers reported in this review. Differences between a marker’s role in rats and humans included where applicable.

	Marker	Biological Role	Localized inflammation/ arthritis	Injury	Infection	Injury + Infection
Cytokines	CINC-1	Cytokine and acute phase mediator59 Inrats: Along with IL-6, induces the synthesis of α2M and AAG59 In humans: Rat CINC-1 is analogous to IL-8 family, specifically melanoma growth-stimulating activity/growth regulated oncogene60,61	? 33,34	-	-	-
	IFN-γ	Multifunctional cytokine;62,63 acute phase mediator;37mainly produced by T cells, B cells, dendritic cells, NK cells, NK T cells, and macrophages62	? 32 *	-	-	? 53
	IL-1β	Pro-inflammatory cytokine;64 acute phase mediator;37 secreted by monocytes and macrophages in response to inflammation64	? 24,32 *	✓ 40	-	? 53
	IL-2	γ-chain cytokine;64 produced by T-cells, dendritic cells, NK cells, and NK T cells64,65	? 32 *	-	-	? 54
	IL-4	γ-chain cytokine;64 produced by TH2 cells, basophils, ILCs, mast cells, and eosinophils64	? 32 *	-	-	? 53
	IL-6	Acute phase mediator;37,64 sources include monocytes, macrophages, endothelial cells, and fibroblasts in response to inflammation64	? 24,32 * ,33,34	? 41	-	? 53
	IL-8	Neutrophil-activating cytokine;64 produced by monocytes, neutrophils, macrophages, lymphocytes, endothelial cells, and epithelial cells in response to IL-1 and TNF-α64,66	? 32 *	-	-	-
	IL-10	Anti-inflammatory cytokine;64,67 secreted by T cells, B cells, NK cells, macrophages, dendritic cells, and monocytes64	? 32 *	? 40	-	? 53
	TGF-β	Multifunctional cytokine;68 acute phase mediator37,68	-	? 40,42	-	-
	TNF-α	Both pro-inflammatory and immunosuppressive cytokine;64 acute phase mediator;37 secreted by monocytes and macrophages in response to tissue injury and infection64,69	? 24	✓ 40	-	✓ 52,53
Acute Phase Proteins	α2M	Protease inhibitor31 In rats: positive acute phase protein31 In humans: not activated in acute phase response70	✓ 31 * ,32 * ,33,34	? 23 * ,31 *	? 23 *	✓ 51
	AAG	Positive acute phase protein;71 functions as a plasma transport protein,72 immune response modulator,73 and binding for drugs73	✓ 29 * ,33,34	-	-	-
	Alb	Negative acute phase protein;37,74 functions to transport substances and maintain osmotic pressure75	? 11,24,30	-	? 11	-
	Cp	Positive acute phase protein that binds to copper76	? 11,30	-	? 11	-
	CRP	Positive acute phase protein37 In humans: roles include complement activation,77 ligand binding,78 and cell-mediated toxicity79	? 10 * ,11,12,22 *	? 41	? 11	✓ 52,53
	Fb	Positive acute phase protein;36 serves as clottingfactor37	✓ 10 * ,11,24,27,35	-	✓ 11	-
	Hp	Positive acute phase protein;80 serves as an anti-inflammatory agent that sequesters free hemoglobin;80 natural bacteriostat81	✓ 6,11,24,27,29 * ,30	-	✓ 11	✓ 48,49
	LCN-2	Positive acute phaseprotein;38 bacteriostatic through sequestering iron82,83	? 38	-	-	-
	SAA	Positive acute phase protein;37 binds cholesterol and modulates immune response72	? 11	-	? 11	-
Other	Cortisol	Glucocorticoid;84 released in response to stress84,85	-	? 41	-	-
	Hb	Plasma protein found in blood following lysis of red blood cells;86,87 circulating source of heme86,87	-	-	-	? 53,55,56
	SAP	Not an acute phase protein in rats or humans39,88	? 22*	-	-	-
	SRM	Mucoprotein fraction of blood serum that contains AAG, alpha-1 cysteine protease inhibitor, and hemopexin89,90	? 29*,30	-	-	-
	WBC	Includes monocytes, lymphocytes, neutrophils, eosinophils, and basophils14	-	? 41	-	? 52,53,55,56

^*indicates articles where statistical methods were not reported

^✓in green denotes markers that have shown changes in response to infection/inflammation and/or markers that displayed similar responses as other longitudinal/terminal measurements (e.g., bacterial counts, imaging scores, etc.)

^?in red denotes markers for which there has been limited testing, or results either are not consistent between models or indicate little to no changes in response to infection/inflammation

^-denotes markers that were not reviewed within a model category

Abbreviations: α2M, α₂-macroglobulin; AAG, α₁-acid glycoprotein; Alb, albumin; CINC, cytokine-induced neutrophil chemoattractant; Cp, ceruloplasmin; CRP, C-reactive protein; Fb, fibrinogen; Hb, hemoglobin; Hp, haptoglobin; IFN, interferon; IL, interleukin; ILC, innate lymphoid cell; LCN, lipocalin; NK, natural killer; SAA, serum amyloid A; SAP, serum amyloid P; SRM, seromucoid; TGF, transforming growth factor; T_H2, T helper 2; TNF, tumor necrosis factor; WBC, white blood cell.

Currently, there is no clear consensus on which blood markers are most sensitive and reliable in tracking the immunological response to inflammation and infection in rats. A multitude of markers have been individually and collectively tested for longitudinal tracking of disease/injury progression with varying success. Notably, the markers explored in this article have been isolated and quantified through a variety of methodologies, including enzyme-linked immunoassays, electrophoresis, and multiplex assays, which may contribute to the variability of reported levels between studies. Additionally, plasma markers have shown potential utility for elucidating the efficacy of treatments for debilitating diseases including arthritis and osteomyelitis in rat models. While the utility of multiplex assays can be limited by the available dynamic range for all factors of interest and the additional assay development/optimization time, separate (singleplex) analysis of multiple markers increases blood volume required. Nonetheless, the use of a variety of markers (acute-phase proteins, pro- and anti-inflammatory cytokines, and WBCs), rather than a single marker, is recommended for the most representative overview of longitudinal immunological response changes in rats. The use of plasma markers as a longitudinal measure of immunological response can accelerate the development and validation of clinically relevant models, accelerating the refinement of these models for enhanced research methodologies and therapeutics.