An investigation and assessment of the muscle damage and inflammation at injection site of aluminum-adjuvanted vaccines in guinea pigs

Abstract

Aluminum salt adjuvants (Als) have been the most widely used adjuvants in vaccines and known to be effective in intramuscular inoculation. However, in rare cases, some Al containing vaccines caused serious adverse events such as chronic pain at the site of the injection. The Als cause mild tissue damage at the inoculation site, allowing the antigen to be locally retained at the inoculation site and thus potentiate innate immunity. This is required to elicit effectiveness of vaccination. However, there is concern that chronic muscle damage might potentially lead to serious adverse events, such as autoimmune disease and movement disorders. In this study, muscle damage caused by several Al containing vaccines were examined in guinea pigs. Mild and moderate inflammation were observed following Al containing split influenza virus vaccine, formalin-inactivated diphtheria-pertussis-tetanus and Salk polio vaccine. While massive inflammation and muscle damage were observed in Al-containing human papillomavirus vaccine-inoculated animals. However, the severities of damage were not associated with their Al contents. Masson’s trichrome staining and immunostaining revealed that injured muscle at the inoculated site recovered within one month of vaccination, whereas inflammatory nodules remained. Flow cytometric analyses of the infiltrating cells revealed that the number of CD45⁺ lymphocytes and potential granulocytes were increased following vaccination. The number of infiltrated cells seemed to be associated with severity of muscle damages. These observations revealed that Al containing vaccine-induced muscle damage is reparable, and severity of transient muscle damages seemed to be determined by type of antigen or types of Al salts rather than Al content.

INTRODUCTION

Aluminum salt adjuvants (Als) as typified by aluminum hydroxide and aluminum phosphate have been widely used as vaccine adjuvant to elicit immunogenicity (Lindblad, 2004), such as formalin-inactivated diphtheria-pertussis-tetanus and conventional Salk inactivated polio vaccine (DPT-cIPV), H5N1 pandemic influenza vaccine, recombinant hepatitis B virus HBV vaccines, hepatitis A vaccine, 7-valent pneumococcal vaccine, and human papillomavirus vaccine (HPVV) in Japan. The Als have been proposed to work as adjuvant by activating innate immunity through the release of damage-associated molecular patterns (DAMPs) from injected tissues by forming weak cytotoxicity (Marichal et al., 2011; Kono and Rock, 2008), and Als can keep the antigen at the injection site (Leeling et al., 1979; Noe et al., 2010; Hansen et al., 2009) to promote its uptake by antigen-presenting cells, such as macrophages and dendritic cells (Morefield et al., 2005; Mannhalter et al., 1985). However, Al-containing vaccines induce some side effects as typified by local swelling, inflammation, and pain, and most of them are mild (Mark et al., 1999; Pittman et al., 2002; Gherardi et al., 1998). Mild muscle tissue damage at the site of inoculation is a common reaction that occurs with vaccines administered intramuscularly (Mrak, 1982; Verdier et al., 2005; Lach and Cupler, 2008). Considering the mechanism of action of Als, i.e., the sustained antigen and DAMP releases, it is reasonable that vaccination into muscular tissues, which feature higher cell density than fatty subcutaneous tissues (Poland et al., 1997), can lead to increased effectiveness of vaccine (Shaw et al., 1989; Groswasser et al., 1997). The appropriate vaccination route should be chosen as the route that induces a sufficient immune response while minimizing local tissue, nerve, and vascular damage. In addition, safety assessment of vaccines is important to evaluate the potential for causing severe adverse events.

Generally, Als have been considered a safe adjuvants for humans; the widespread use of Al adjuvant can be attributed in part to their excellent safety record based on a 70-year history of use (Hogenesch, 2013). However, in rare cases, some Al-containing vaccines, such as HPVV, have been reported to cause serious adverse reactions, such as chronic pain at the site of the injection, unlike other Al-containing vaccines (Miranda et al., 2017; Inbar et al., 2017; Arnheim-Dahlström et al., 2013; Palmieri et al., 2017; Ozawa et al., 2017; Tomljenovic and Shaw, 2012). It has been unclear whether the Al or antigen is involved in this reaction. It has been speculated that such severe adverse events might be caused by autoimmune-mediated mechanisms (Palmieri et al., 2017; Arnheim-Dahlström et al., 2013; Miranda et al., 2017). Als have the potential to trigger inflammation and retain antigen immunogenicity at injection sites. It has been reported that chronic immune activation can lead to an increased risk of autoimmune disease development (Ishihara and Hirano, 2002; Murakami and Hirano, 2012). Thus, assessment of severity and chronical damage of injection sites is important to understanding Al-containing vaccine-induced adverse reactions.

There have been few reports on evaluation of muscle injury at the injection site of Al-containing vaccines in animal models (Kashiwagi et al., 2014; Weeratna et al., 2000). In addition, there have been no reports focused on the reversibility of muscle damage from HPVV vaccination or the comparison of the degree of muscle damage between different Al-containing vaccines.

In this study, we assessed intramuscular vaccination-induced muscle damage using HPVV (virosomal antigen + Al), DPT-cIPV (whole cell/particle antigen + Al), and the Al-adjuvanted trivalent inactivated influenza vaccine (Alum-TIV, split antigen + Al). The incidence of adverse events of these vaccines is summarized in Table 1 (The Vaccine Adverse Reactions Review Committee of Japan; U.S. Department of Health and Human Services, The Vaccine Adverse Event Reporting System). Using these vaccines, we evaluated muscle damage and recovery at the site of intramuscular inoculation. We also performed flow cytometric analyses of the infiltrating cells at the inoculation site in muscle tissues to explore the possible associations between the immune and inflammatory responses at the injection site. Based on the results, we discuss impact of Als and the antigen on severity of muscle damage and its recovery.

Table 1. Incidence of adverse events in evaluated vaccines.

Vaccine	Adverse event	Country	Calculate	2012	2013	2014	2015	2016
DPT-IPV	Pain (not limited to injection sites)	Japan	Absolute number	0	0	1	0	0
			Rate of cases/per 100,000 shots	0	0	0.116	0	0
		USA	Absolute number	8	5	17	10	16
			Rate of cases/per 100,000 shots	0.0654	0.0406	0.136	0.0793	0.126
	Adverse events at injection sites*	Japan	Absolute number	1	2	1	0	0
			Rate of cases/per 100,000 shots	0.0355	0.109	0.116	0	0
		USA	Absolute number	21	25	35	23	28
			Rate of cases/per 100,000 shots	0.172	0.203	0.281	0.182	0.221
Influenza vaccine	Pain (not limited to injection sites)	Japan	Absolute number	22	47	27	60	52
			Rate of cases/per 100,000 shots	0.0438	0.0909	0.0515	0.117	0.107
		USA	Absolute number	3310	3510	3700	3590	2810
			Rate of cases/per 100,000 shots	3.74	3.62	3.68	3.47	2.6
	Adverse events at injection sites*	Japan	Absolute number	22	30	30	39	49
			Rate of cases/per 100,000 shots	0.0438	0.058	0.0573	0.0758	0.101
		USA	Absolute number	2250	2470	2550	2570	2290
			Rate of cases/per 100,000 shots	2.54	2.55	2.54	2.49	2.11
HPVV**	Pain (not limited to injection sites)	Japan	Absolute number	238	220	46	40	1
			Rate of cases/per 100,000 shots	9.42	35.9	94.99	153.85	8.57
		USA	Absolute number	364	393	335	380	470
			Rate of cases/per 100,000 shots	15.39	13.13	8.8	8.73	10.08
	Adverse events at injection sites*	Japan	Absolute number	33	26	1***	1***	0***
			Rate of cases/per 100,000 shots	1.31	4.24	2.07***	3.85***	0***
		USA	Absolute number	270	240	243	255	316
			Rate of cases/per 100,000 shots	11.42	8.02	6.38	5.86	6.78

*Adverse events at injection sites include swelling, pain, and redness.

**HPVV includes Cervarix (2-valent vaccine) and Gardasil (4-valent vaccine) and the latter is used in the present study.

***Total shot number is below 30,000 shots.

MATERIALS AND METHODS

Animals and ethics statement

Female Hartley guinea pigs (280–300 g body weight, SPF grade) were purchased from SLC (Shizuoka, Japan). They were housed in solid-bottomed metal cages and fed on standard guinea pig chow. Chow and water were provided ad libitum. The temperature was maintained at 24–26°C with a 12-hr light/dark cycle. Before the experiments, the animals were acclimated for 2 weeks. All animal experiments were performed according to the guidelines of the Ethics Review Committee of Animal Experiments at the National Institute of Infectious Diseases in Japan (approval numbers 715030 in 2015-2017).

Vaccination procedures and experimental design

The guinea pigs received one or three injections (1 week intervals) of saline (SA); HPVV containing aluminum hydroxyphosphate sulfate (225 µg/0.5 mL as aluminum), human papillomavirus 1 type 6 L1 protein (20 µg/0.5 mL), human papillomavirus 1 type 11 L1 protein (40 µg/0.5 mL), human papillomavirus 1 type 16 L1 protein (40 µg/0.5 mL), and human papillomavirus 1 type 18 L1 protein (20 µg/0.5 mL) (Gardasil, MSD K.K, Tokyo, Japan); DPT-cIPV containing aluminum (III) chloride hexahydrate (100.65 µg/0.5 mL as aluminum), ≥ 4 units/0.5 mL of the Bordetella pertussis protective antigen, ≤ 15 (Limit of Flocculation, Lf) Lf/0.5 mL (≥ 14 international units) of diphtheria toxoid, ≤ 2.5 Lf/0.5 mL (≥ 9 international units) of tetanus toxoid, 40 D antigen unit (DU)/0.5 mL of inactivated poliovirus type 1, 8 DU/0.5 mL of inactivated poliovirus type 2, and 32 DU/0.5 mL of inactivated poliovirus type 3 (Daiichi Sankyo, Tokyo, Japan, formerly Kitasato Daiichi Sankyo Co., Ltd, Tokyo, Japan); or in house Alum-TIV containing aluminum hydroxide (500 µg /0.5 mL as aluminum, Alhydrogel adjuvant, InvivoGen, San Diego, California, USA) and 15 µg/0.5 mL of hemagglutinin per virus strain (A/California/7/2009 (H1N1), A/New York/39/2012 (H3N2), and B/Massachusetts/2/2012 strains) (Chemo-Sero-Therapeutic Research Institute, Kaketsuken, Kumamoto, Japan). All commercially available vaccines were used before the expiry date. The volume of 0.5 mL per guinea pig of the vaccine or SA is equivalent to the human inoculum. The inoculation site was the left quadricep muscle. For the single vaccination studies, 48 hr post-vaccination, the animals were anesthetized by intraperitoneal injection of 10 mg/kg of pentobarbital sodium. The blood was immediately collected by cardiac puncture and added to an EDTA-coated tube. The blood was used for white blood cell (WBC) counts and to obtain the serum. Serum was obtained using a Capiject (Terumo, Tokyo, Japan). The quadricep was removed and placed on ice in a 5% FCS-supplemented PBS solution before being used for flow cytometric analyses. For the repeated vaccination studies, the inoculation site was the same in the previously injected left quadricep muscle and at 48 hr post-final vaccination, the animals were sacrificed and processed following the same scheme as used for the single vaccination studies described above. For the recovery studies, animals were sacrificed and sampled 1 month after the final vaccination. For the repeat vaccination study, the animal’s body weight was monitored at days 0 (immediately before vaccination), 1, 2, 3, 7, 14, and 44.

Blood biochemistry tests and white blood cell counting

The WBC counts were determined using an automatic blood cell analyzer, the MEK-6450 (Nihon Koden, Tokyo, Japan). The plasma blood urea nitrogen (BUN), creatinine (Cre), creatinine kinase (CK), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (T-Bil), lactate dehydrogenase (LDH), alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), and albumin (ALB) levels were measured using DRI-CHEM (Fujifilm, Tokyo, Japan).

Histopathological analyses

Quadricep samples from the vaccine inoculation site were excised and fixed in 10% neutral-buffered formalin. The fixed samples were dehydrated through graded alcohol washes and embedded in paraffin. Serial sections of 4 µm thickness were stained with hematoxylin and eosin (H&E) and observed under a microscope for routine pathological examination. To assess the skeletal muscle fiber architecture, paraffin-embedded sections were stained with Masson’s trichrome using a commercial kit (Sigma-Aldrich, St. Louis, MO, USA). Immunohistochemistry for the detection of desmin was performed with sections incubated overnight at 4°C with the rabbit monoclonal anti-desmin antibody (clone Y66, Abcam, Cambridge, UK, 1:50). The anti-rabbit IgG secondary antibody with fluorescein isothiocyanate conjugate (1:200) was incubated for 1 hr at room temperature. Images were acquired on an Olympus BX53 (Tokyo, Japan).

Immune infiltrate analysis using a fluorescence-activated cell sorting system

The single cell preparations from the vaccine-inoculated quadricep of each animal were obtained based on a previously published protocol (Tierney et al., 2014) with slight modifications. Briefly, single cell suspensions were obtained by enzymatic digestion of muscles with collagenase type IV (0.5 mg/mL, Sigma Aldrich) and dispase (3.5 mg/mL, Invitrogen) in HEPES-buffered RPMI 1640 at 37°C for 40 min. The digestion reactions were stopped by adding 10 mM EDTA and placing the samples on ice. The cells were passed through a 70 µm filter and collected. Approximately 1.0 × 10⁶ cells were incubated for 30 minutes in PBS containing 5% FCS and 0.1 mM EDTA with the fluorescein isothiocyanate-conjugated CD25 antibody (BD, clone 7D4, 1:100) and the phycoerythrin-conjugated CD8 antibody (BD, clone 53–6.7, 1:50). Propidium iodide solution (PI, BD Biosciences, 10 µL/1.0 × 10⁶ cells) was used to differentiate live and dead cells. The cells were analyzed with a CytoFLEX (Beckman-Coulter, Pasadena, CA, USA). Acquired data were analyzed with the FlowJo 10.1r5 (Treestar, LLC, USA).

Statistics

Statistical analyses were performed using GraphPad Prism software version 6.0 (GraphPad Software, Inc., San Diego, CA, USA). Statistical significance was determined with the one-way ANOVA test followed by Dunnett’s multiple-comparison test. Differences were considered to be significant at P < 0.05. Data are presented as the mean ± SD.

RESULTS

Changes in body weight and WBC composition

The present study used the guinea pig as an experimental animal model because guinea pigs have been widely used in the lot release safety tests, such as the abnormal toxicity test or general safety test for vaccines in many countries as well as Japan (Mizukami et al., 2009). The types of Al and its content in tested vaccine are summarized in Table 2.

Table 2. Content of aluminum salt adjuvant in vaccines.

Vaccine	Manufacturer	Vaccination route in humans	Aluminum adjuvant	Aluminum content	Aluminum content/dose
DPT-cIPV	Kitasato Institute	Subcutaneous injection	Al(Cl)3·6H2O	201.3 µg/mL	100.65 µg
Alum-TIV	Invivogen for Al; Chemo-Sero-Therapeutic Research Institute for TIV	Subcutaneous injection (without Al)	Al(OH)3	1000 µg/mL	500 µg
HPVV	MSD	Intramuscular injection	Al2(PO4)(OH)(SO4)	450 µg/mL	225 µg

There were no deaths and no remarkable clinical signs observed throughout the study. The body weight changes after vaccination were monitored at days 0, 1, 2, 3, 7, 14, and 44. For the repeat vaccination study, second and third vaccinations were conducted on day 7 and 14, respectively. The results showing that no significant changes occurred among SA-, Alum-TIV, HPVV- and DPT-cIPV-injected animals (Fig. 1A). These results suggest that remarkable systematic toxic reactions did not occur in the present vaccination study. Next, to verify whether systemic inflammatory reactions were occurred, the number of WBCs in peripheral blood was measured 48 hr after the vaccination. The number of WBC were not altered by single inoculation with any of vaccines (Fig. 1B). However, the repeat injection study showed a significant increase in the number of WBCs in the HPVV-injected group (Fig. 1B); DPT-cIPV- or Alum-TIV-treated animals did not show changes in the number of WBC (Fig. 1B). One month after the last HPVV vaccination, the number of WBCs recovered to normal levels; the number of WBCs was approximately equivalent to the number found in the SA-treated animals (Fig. 1B). These results suggest that temporary intense inflammatory reactions occurred in animals that received repeat HPVV vaccination.

Fig. 1

Time-dependent changes in body weight, white blood cell (WBC) count, and blood biochemistry. Guinea pigs were intramuscularly injected with vaccine into the left quadricep muscle once (single injection study) or three times (repeated injection study). The body weight was monitored on days 0, 7, 14, 16, and 44 (A). The WBC count was analyzed 48 hr or 1 month after the single or the third of three vaccination(s) (B). The blood biochemical parameters were assessed 48 hr or 1 month after the single or the third of three vaccination(s) (C–E). The data are displayed as the mean ± SD (A); each plot indicates the individuals, and the error bar indicates ± SD (B–E). * P < 0.05; ** P < 0.01.

Blood biochemistry

To evaluate the general toxic effects of the vaccinations on major organs, blood biochemical analyses were performed following repeated vaccination with SA, DPT-cIPV, Alum-TIV or HPVV (Fig. 1C–E). The serum Cre and BUN levels, which are biomarkers for kidney failure (Gowda et al., 2010), were largely unchanged in the vaccinated groups (Fig. 1C). Lower levels of BUN were observed 48 hr after the third immunization in the HPVV-vaccinated animals compared with SA-treated animals.

The serum levels of AST, CPK, and LDH were measured; these enzymes may act as an index of cellular necrosis or tissue damage following acute and chronic tissue injuries (Fig. 1D). The results indicate that the animals treated with 3 times repeated vaccination of the HPVV had increased levels of these enzymes compared with the SA-treated animals (Fig. 1D). However, the increases in AST, CPK, and LDH levels were not observed 1 month after the last HPVV vaccination (Fig. 1D). In addition, animals treated with 3 times repeated vaccination of the DPT-cIPV did not show an increase in these enzyme levels compared with SA-treated animals (Fig. 1D). These results suggest that repeated intramuscular injection of the HPVV caused cellular necrosis or tissue damage at 48 hr post-vaccination (Fig. 1D). However, cellular necrosis or tissue damage might be recovered within 1 month post-vaccination (Fig. 1D).

To assess the liver function following the vaccinations, serum ALT, T-Bil, GGT, and ALB levels were measured (Fig. 1E). Although mild changes in the levels of these biomarkers were observed, no remarkable changes were identified. In addition, these changes are within the range of normal values in guinea pigs based on reference data (Clampitt and Hart, 1978; Clifford and White, 1999; Maureen and Laura, 2012; Rabe, 2011). Taken together, these results suggest that hepatic injury did not occur with the present intramuscular vaccination method.

Histopathological changes in the inoculation site in a single vaccination study

Blood biochemical analyses suggested that animals injected intramuscularly with the HPVV suffered remarkable tissue damages or organ injuries that could recover 1 month after inoculation. To understand this phenomenon, we performed histopathological analyses focused on the vaccine injection site at 48 hr after vaccination in a single injection study (Fig. 2A).

Fig. 2

Hematoxylin and eosin (H&E) staining of skeletal muscle at the injection site. Guinea pigs were intramuscularly injected with vaccine into the left quadricep muscle once (single injection study) or three times (repeated injection study). The injection sites were removed and fixed with formalin, then embedded in paraffin. Sections were stained with H&E and observed under a microscope for routine pathological examination. The top (A) and bottom panel (B) indicate the results following single injection and three repeated injections, respectively. Staining was performed 48 hr after vaccination, and sections from one month after the last vaccination were used for staining in HPVV-treated animals. Each panel indicates the result from an individual. Data are shown at 2 × magnification.

Histopathological changes at the injection site are shown in Fig. 2A. Almost no or mild changes are seen in the SA-treated animals and consist of myofiber necrosis, inflammation, fibroblast proliferation in the muscle, and hemorrhage in the interstitial connective tissue (Fig. 2A). The inflammatory cells were speculated to be macrophages and neutrophils. DPT-cIPV-treated or Alum-TIV-treated animals showed localized myofiber necrosis and migration of inflammatory cells at the injection site, while HPVV-treated animals showed extensive necrosis and infiltration of inflammatory cells (Fig. 2A). The area and severity of necrosis and inflammatory cell infiltration seem to not correlate with the Al content of the vaccine (Table 2); Al content is highest in the Alum-TIV and lowest in the DPT-cIPV (Alum-TIV Al content > HPVV Al content > DPT-cIPV Al content), as shown in Table 2. This result suggests that severity of muscle damages does not depend on Al content.

Histopathological changes in the inoculation site in a repeated vaccination study

Similarly, we performed histopathological analyses focused on the vaccine injection site 48 hr after the last vaccination (Fig. 2B). Almost no or mild changes were seen in the SA-treated animals and consisted of myofiber necrosis, inflammation, fibroblast proliferation in the muscle, and hemorrhage in the interstitial connective tissue (Fig. 2B). DPT-cIPV- or Alum-TIV-treated animals showed local myofiber necrosis, migration of inflammatory cells, and fibroblast proliferation at the injection site, while HPVV-treated animals showed more extensive necrosis, infiltration of inflammatory cells, and fibroblast proliferation (Fig. 2B). To investigate whether muscle damage following the HPVV vaccination is a chronic or recovery phenomenon, the inoculation sites were analyzed one month after the last HPVV vaccination. Although muscle fiber heterogeneities were partially observed, noticeable muscle fiber necrosis, fibrotic changes, and fibroblast proliferation were not found (Fig. 2B). These results suggest that repeated intramuscular injection at one-week intervals causes more severe myofiber necrosis or abundant immune cell infiltration than a single injection. Although the HPVV induced the most severe myofiber necrosis among the tested vaccines, the disruption of muscle fibers was repaired within one month of the last vaccination. These results suggest that HPVV-induced severe muscle damages were not chronic and were reparable.

Regeneration of skeletal muscle during recovery from vaccination-induced muscle damage

Masson’s trichrome staining was performed to verify the recovery from muscle damage following repeated vaccination of Al-containing vaccines (Fig. 3A–E). At 48 hr after the last vaccination, widespread damage was observed in the HPVV-treated animals (Fig. 3D). In the same manner, the DPT-cIPV- or Alum-TIV-treated animals showed a wide range of tissue disruption, although the muscle damages were less notable in these animals than in the HPVV-treated animals (Fig. 3B and D). One month post-vaccination, extensive regeneration had occurred in the HPVV-treated animals (Fig. 3E). However, inflammatory nodules remained (Fig. 3E). This reveals that muscle damage due to the HPVV intramuscular administration was transient, and the disrupted muscle was repaired via probably myofibroblast and myoblast activation.

Fig. 3

Masson’s trichrome staining of vaccine-injected skeletal muscle. Guinea pigs were intramuscularly injected with vaccine into the left quadricep muscle once (single injection study) or three times (repeated injection study). The injection sites were removed and fixed with formalin, then embedded in paraffin. Sections were stained with the Masson’s trichrome staining kit and observed under a microscope. Staining was performed 48 hr after vaccination with (A) saline (SA), (B) diphtheria-pertussis-tetanus and formalin-inactivated Salk polio vaccine (DPT-cIPV), (C) aluminum-containing hemagglutinin split influenza vaccine (Alum-TIV), or (D) human papillomavirus vaccine (HPVV). Staining was performed one month after vaccination with the HPVV (E). Each panel indicates a result from an individual (n = 3). Data are shown at 4 × magnification.

Desmin is an intermediate filament protein that is highly expressed during early muscle differentiation; its expression continues into the mature adult myofiber (Costa et al., 2004). Using immunohistochemistry, we analyzed desmin expression to further evaluate the regeneration of muscle fibers following the HPVV-induced muscle damage (Fig. 4A–E). At 48 hr after the last vaccination, desmin expression was visibly reduced in the HPVV-treated animals (Fig. 4D). In addition, a decrease in desmin expression was observed in the DPT-cIPV- and Alum-TIV-treated animals; this decrease was visibly most severe in the HPVV-treated animals, followed by the DPT- and then the Alum-TIV-treated animals (Fig. 4B–D). One month post-vaccination, desmin expression recovered in the HPVV-treated animals (Fig. 4E) to the level found in the SA-treated animals (Fig. 4A); the recovery was homogeneous throughout 3 individual animals. This suggests that muscle damage following intramuscular injection of the HPVV is almost completely healed by muscle regeneration one month after inoculation.

Fig. 4

Desmin expression in regenerating vaccine-injected skeletal muscle. Guinea pigs were intramuscularly injected with vaccine into the left quadricep muscle once (single injection study) or three times (repeated injection study). The injection sites were removed and fixed with formalin, then embedded in paraffin. Sections were stained for desmin and observed under a fluorescent microscope. Staining was performed 48 hr after vaccination with (A) saline (SA), (B) diphtheria-pertussis-tetanus and formalin-inactivated Salk polio vaccine (DPT-cIPV), (C) aluminum-containing hemagglutinin split influenza vaccine (Alum-TIV), or (D) human papillomavirus vaccine (HPVV). Staining was performed one month after vaccination with the HPVV (E). Each panel indicates a result from an individual (n = 3). Data are shown at 10 × magnification.

Assessment of lymphocyte and granulocyte infiltration in injured muscles

To quantify the involvement of the immune system in the immediate response to skeletal muscle damage caused by intramuscular vaccination, we analyzed the number of granulocytes and mononuclear cells at the injection site at 48 hr after vaccination. Flow cytometric analyses were performed using SSC, FSC, anti-CD45 antibodies, and anti-CD18 antibodies; these parameters were selected due to limitations in the availability of antibodies with cross-reactivity to guinea pig antigens. Following single or repeated injections, we collected and digested the muscle and isolated the infiltrating granulocytes and mononuclear cells for characterization. Relatively few granulocytes and CD45⁺ cells were detected in the single inoculation SA- or DPT-cIPV-treated animals (Fig. 5A). However, following the single inoculation with HPVV, the number of CD45⁺ leukocytes was significantly increased (Fig. 5A).

Fig. 5

Flow cytometric analysis of infiltrated cells in vaccine-injected skeletal muscle. Guinea pigs were intramuscularly injected with vaccine into the left quadricep muscle once (A) or three times (B). At 48 hr after vaccination, the quadricep muscles were collected. The single cell suspensions were prepared and stained with anti-CD45 and anti-CD18 antibodies. The left panels indicate the total lymphocyte and macrophage populations and the CD45+ population; the right panels indicate dot plots for the SSC vs. FSC gate and CD45 vs. CD18 gate. Cells were collected following single vaccination (A) or three repeated vaccinations (B). Each plot indicates an individual, and the error bar indicates ± SD (left panels).

Following repeated vaccinations, the number of granulocytes and CD45⁺ cells significantly increased in the HPVV-treated animals (Fig. 5B). The DPT-cIPV-treated animals also showed significant increases in the number of CD45⁺ cells compared with the SA-treated animals (Fig. 5B). CD18 has been known to act as the beta-subunit for different structures, as typified by integrin alphaX/beta2. Integrins are known to participate in cell adhesion as well as cell surface-mediated signaling (Gjelstrup et al., 2010). However, no CD18⁺ cells were observed in any of the vaccinated individuals (Fig. 5A and B). These results indicate that muscle caused by intramuscular vaccination was associated with immune cell infiltrations. However, CD11b⁺ or CD11c⁺ cells, such as dendritic cells or macrophages, were not a major component of the infiltrated cells, because most of dendritic cells or macrophages express CD18⁺ antigen (Gjelstrup et al., 2010).

DISCUSSION

In this study, we assessed severity of muscle damage on intramuscular injection of approved Al-containing vaccines using guinea pigs. The results of blood biochemical analyses suggest that remarkable tissue damages or organ injuries were occurred following vaccination with the HPVV (Fig. 1C). However, biomarker of liver injury and kidney injury were not notably elevated (Figs. 1C and 1E).

Histopathological analyses revealed that severity of muscle damages (necrosis and inflammation) was more pronounced after injection of the HPVV than with the other Al-containing vaccines (Fig. 2A and B). The Al content of the DPT-cIPV, Alum-TIV, and HPVV was 201.3 µg/mL, 1000 µg/mL, and 450 µg/mL, respectively. Although damage in muscle fibers and infiltration of immune-related cells occurred following all the vaccinations tested, the severity of necrotic changes in the muscle fibers and infiltration of immune-related cells did not correlate with aluminum content (Figs. 2–4). One possible explanation for this discordance is that muscle damage and inflammatory reactions may be affected by not only Al content but also by the immunogenicity of the antigen. The HPVV is composed of virus-like particles (VLP) which have a structure mimicking whole virus particles; the antigenicity has been considered to be higher than that of split-type vaccines (Chroboczek et al., 2014). The immunogenicity of the TIV, which consists predominantly of purified hemagglutinin proteins, is known to be lower than that of whole particle or virosomal influenza vaccines (Soema et al., 2015). This supports the idea that muscle damage caused by intramuscular vaccination is also affected by the antigen’s immunogenicity. In addition, notably, the types of salts of Al containing tested vaccines used are different (Table 2). The aluminum hydroxyphosphate sulfate is known to have a higher ability to enhance the effectiveness of the HPVV than a conventional aluminum hydroxide adjuvant (Caulfield et al., 2007). Conventional aluminum hydroxide adjuvant is supplemented in the Alum-TIV and DPT-cIPV that were used in this study. Therefore, although Al has cytotoxic effects and enhances immune cell accumulation at the injection site, the severity of muscle injuries may also depend in part on the properties of the salt type of Al. To evaluate the differences of the salt type of Al on muscle damage, it is necessary to evaluate muscle damage with the Al adjuvant alone. However, aluminum hydroxyphosphate sulfate could not be commercially obtained and therefore investigation was limited. It is a future issue to assess and compare muscular damage by different salt types of Al.

The incidence of adverse events of the evaluated vaccines in Japan and the USA is summarized in Table 1. The references for the incidence of adverse events in Japan and the USA are The Vaccine Adverse Reactions Review Committee of Japan and The Vaccine Adverse Event Reporting System, respectively (The Vaccine Adverse Reactions Review Committee of Japan; U.S. Department of Health and Human Services, The Vaccine Adverse Event Reporting System). As a point of caution, in Table 1, the incidences in HPVV include both Cervarix (2-valent vaccine) and Gardasil (4-valent vaccine), and the latter was used present study. Table 1 shows that HPVV has a relatively high incidence for pain and adverse events at injection sites compared with other two vaccines. The incidence of these side effects in humans seems to be likely to correlate with the degree of muscle injury and inflammation at the injection site in guinea pigs (Figs. 2–4). These data suggest that evaluation of the injection site in guinea pigs may be useful in predicting adverse reactions in injection site in humans.

Unlike the myocardium, skeletal muscle recovers from damage with myofibroblast activation (Mann et al., 2011). When a compound having cytotoxicity, like an aluminum-adjuvanted vaccine, persistently exists in the injured tissue, chronic immune activation can occurr and this can lead to an increased risk of autoimmune disease development (Ishihara and Hirano, 2002; Murakami and Hirano, 2012). The Al acts to keep the antigen at the site of inoculation, which may prolong the immune response at the injection site (Leeling et al., 1979; Noe et al., 2010; Hansen et al., 2009; Manhalter et al., 1985). Kashiwagi et al. (2014) reported that Al containing vaccines induce inflammatory nodules at the injection site, and that the inflammatory nodules remain 6 months later. In this study, we examined the recovery from the severe muscle damage caused by the HPVV vaccination. The damaged muscles were reduced within 1 month after vaccination (Fig. 2B). However, the results of Masson’s trichrome staining revealed that fibrotic tissues and inflammatory nodules remained one month after vaccination (Fig. 3E). Immunostaining with the anti-desmin antibody was carried out to verify the integrity of the muscle fibers in the tissues where recovery was observed with H&E staining. One month after inoculation, the desmin staining intensity was nearly identical in the HPVV-treated and SA-treated animals (Fig. 4E). This suggests that severe muscle damages were recovered 1 month after the last vaccination. However, inflammatory nodules were still observed at the injection site (Fig. 3E). It has been reported that chronic inflammatory reactions in damaged muscles are involved in the development of autoimmune responses (Ishihara and Hirano, 2002; Murakami and Hirano, 2012). However, since the inoculation regimen (vaccination interval) in this study is different from that of humans, there is a limitation to consider regarding the responses in clinical usage of the evaluated vaccines. Future studies are needed to overcome this limitation and demonstrate the hypothesis that chronic inflammatory reaction might be involved in reactogenicity of HPVV in humans.

Damage of muscle fibers may occur due to the physical destruction caused by the injection of the vaccine or due to the inflammatory reaction caused by granulocytes drawn to the injection site by the vaccine adjuvant and antigen (Lu and Hogenesch, 2013). Thus, both physical destruction and immune-mediated reactions were considered to be involved in the observed muscle damages following Al-containing vaccine inoculations. The result of flow cytometric analyses revealed that the HPVV and DPT-cIPV vaccinations resulted in an increase in the number of granulocytes and CD45⁺ cells (Fig. 5A and B). This suggests that the inflammatory response by granulocytes and lymphocyte might be accelerating both muscle fiber destruction and repair as previously reported (Chazaud et al., 2009).

The rapid resolution of tissue damages requires a sequential and well-orchestrated series of events. Perturbation of any of these stages can result in unsuccessful muscle regeneration, typically characterized by the persistent degeneration of myofibers, inflammation, and fibrosis (Gupta et al., 2005; Kääriäinen et al., 2000; Wynn, 2008). Immediately after occurring skeletal muscle damages, cytokines and growth factors are released from infiltrating inflammatory cells (Chazaud et al., 2009). These factors stimulate the migration of the inflammatory cells to and at the site of injured and mediate satellite cell proliferation and cell survival. In addition, the importance of T-cell-produced cytokines in both muscle degeneration and repair has been reported (Lluís et al., 2001; Farini et al., 2007; Vetrone et al., 2009). However, since commercially available antibodies that react to lymphocytes in guinea pigs are limited, we could not observe further analyze the subsets of CD45⁺ cells.

Intramuscular injection is an effective method of vaccination; however, persistent inflammation is often induced, leading to an increased risk of severe and chronic side effects. Persistent inflammation can also trigger chronic immune reactions (Ishihara and Hirano, 2002; Murakami and Hirano, 2012). In rare cases, HPVV has been reported to caused chronic pain in injection cites and severe consciousness disorder (Arnheim-Dahlström et al., 2013; Ozawa et al., 2017). These reactions have been considered to be caused by immune-mediated mechanisms. However, the influence of transient muscle damages in the inoculation site on the development of autoimmune reactions following HPVV inoculation in humans remains unknown. Therefore, further studies are needed to link local transient muscle damages and HPVV-related severe adverse events in humans.

In summary, we analyzed the toxicity of Al containing vaccine by focusing on the inoculation site in guinea pigs. The severity of necrosis and inflammatory responses at the inoculated site did not depend on the vaccine’s Al content. In addition, although severe inflammation and necrosis were observed with the HPVV, most muscle fibers recovered one month post-vaccination with remaining inflammatory nodules. Notably it is still unknown whether severity of transient muscle damages is associated with Al-containing vaccine-related severe adverse events in humans. To clarify this issue, further studies are needed to elucidate the mechanism of Al containing vaccine-induced innate and adaptive immune responses in humans.

ACKNOWLEDGMENTS

The authors would like to thank Dr. Iwao Kukimoto (Pathogen Genomics Center, National Institute of Infectious Diseases, Tokyo, Japan) and Dr. Keiko Tanaka-Taya (Center for Surveillance, Immunization, and Epidemiologic Research, National Institute of Infectious Diseases, Tokyo, Japan) for helpful discussion and advice.

Conflict of interest

The authors declare that there is no conflict of interest.

REFERENCES

•  Arnheim-Dahlström,  L.,  Pasternak,  B.,  Svanström,  H.,  Sparén,  P. and  Hviid,  A. (2013): Autoimmune, neurological, and venous thromboembolic adverse events after immunisation of adolescent girls with quadrivalent human papillomavirus vaccine in Denmark and Sweden: cohort study. BMJ, 347, f5906.
•  Caulfield,  M.J.,  Shi,  L.,  Wang,  S.,  Wang,  B.,  Tobery,  T.W.,  Mach,  H.,  Ahl,  P.L.,  Cannon,  J.L.,  Cook,  J.C.,  Heinrichs,  J.H. and  Sitrin,  R.D. (2007): Effect of alternative aluminum adjuvants on the absorption and immunogenicity of HPV16 L1 VLPs in mice. Hum. Vaccin., 3, 139-145.
•  Chazaud,  B.,  Brigitte,  M.,  Yacoub-Youssef,  H.,  Arnold,  L.,  Gherardi,  R.,  Sonnet,  C.,  Lafuste,  P. and  Chretien,  F. (2009): Dual and beneficial roles of macrophages during skeletal muscle regeneration. Exerc. Sport Sci. Rev., 37, 18-22.
•  Chroboczek,  J.,  Szurgot,  I. and  Szolajska,  E. (2014): Virus-like particles as vaccine. Acta Biochim. Pol., 61, 531-539.
•  Clampitt,  R.B. and  Hart,  R.J. (1978): The tissue activities of some diagnostic enzymes in ten mammalian species. J. Comp. Pathol., 88, 607-621.
• Clifford, C.B. and White, W.J. (1999): The guinea pig. In: The clinical chemistry of laboratory animals (Loeb, W.F., Quimby, F.W., ed.), second ed., pp. 65-70, CRC Press, Florida.
•  Costa,  M.L.,  Escaleira,  R.,  Cataldo,  A.,  Oliveira,  F. and  Mermelstein,  C.S. (2004): Desmin: molecular interactions and putative functions of the muscle intermediate filament protein. Braz. J. Med. Biol. Res., 37, 1819-1830.
•  Farini,  A.,  Meregalli,  M.,  Belicchi,  M.,  Battistelli,  M.,  Parolini,  D.,  D’Antona,  G.,  Gavina,  M.,  Ottoboni,  L.,  Constantin,  G.,  Bottinelli,  R. and  Torrente,  Y. (2007): T and B lymphocyte depletion has a marked effect on the fibrosis of dystrophic skeletal muscles in the scid/mdx mouse. J. Pathol., 213, 229-238.
•  Gherardi,  R.K.,  Coquet,  M.,  Chérin,  P.,  Authier,  F.J.,  Laforêt,  P.,  Bélec,  L.,  Figarella-Branger,  D.,  Mussini,  J.M.,  Pellissier,  J.F. and  Fardeau,  M. (1998): Macrophagic myofasciitis: an emerging entity. Groupe d’Etudes et Recherche sur les Maladies Musculaires Acquises et Dysimmunitaires (GERMMAD) de l’Association Française contre les Myopathies (AFM). Lancet, 352, 347-352.
•  Gjelstrup,  L.C.,  Boesen,  T.,  Kragstrup,  T.W.,  Jørgensen,  A.,  Klein,  N.J.,  Thiel,  S.,  Deleuran,  B.W. and  Vorup-Jensen,  T. (2010): Shedding of large functionally active CD11/CD18 Integrin complexes from leukocyte membranes during synovial inflammation distinguishes three types of arthritis through differential epitope exposure. J. Immunol., 185, 4154-4168.
•  Gowda,  S.,  Desai,  P.B.,  Kulkarni,  S.S.,  Hull,  V.V.,  Math,  A.A. and  Vernekar,  S.N. (2010): Markers of renal function tests. N. Am. J. Med. Sci., 2, 170-173.
•  Groswasser,  J.,  Kahn,  A.,  Bouche,  B.,  Hanquinet,  S.,  Perlmuter,  N. and  Hessel,  L. (1997): Needle length and injection technique for efficient intramuscular vaccine delivery in infants and children evaluated through an ultrasonographic determination of subcutaneous and muscle layer thickness. Pediatrics, 100, 400-403.
•  Gupta,  S.,  Gallavotti,  A.,  Stryker,  G.A.,  Schmidt,  R.J. and  Lal,  S.K. (2005): A novel class of Helitron-related transposable elements in maize contain portions of multiple pseudogenes. Plant Mol. Biol., 57, 115-127.
•  Hansen,  B.,  Belfast,  M.,  Soung,  G.,  Song,  L.,  Egan,  P.M.,  Capen,  R.,  Hogenesch,  H.,  Mancinelli,  R. and  Hem,  S.L. (2009): Effect of the strength of adsorption of hepatitis B surface antigen to aluminum hydroxide adjuvant on the immune response. Vaccine, 27, 888-892.
•  Hogenesch,  H. (2013): Mechanism of immunopotentiation and safety of aluminum adjuvants. Front. Immunol., 3, 406.
•  Inbar,  R.,  Weiss,  R.,  Tomljenovic,  L.,  Arango,  M.T.,  Deri,  Y.,  Shaw,  C.A.,  Chapman,  J.,  Blank,  M. and  Shoenfeld,  Y. (2017): Behavioral abnormalities in female mice following administration of aluminum adjuvants and the human papillomavirus (HPV) vaccine Gardasil. Immunol. Res., 65, 136-149.
•  Ishihara,  K. and  Hirano,  T. (2002): IL-6 in autoimmune disease and chronic inflammatory proliferative disease. Cytokine Growth Factor Rev., 13, 357-368.
•  Kashiwagi,  Y.,  Maeda,  M.,  Kawashima,  H. and  Nakayama,  T. (2014): Inflammatory responses following intramuscular and subcutaneous immunization with aluminum-adjuvanted or non-adjuvanted vaccines. Vaccine, 32, 3393-3401.
•  Kono,  H. and  Rock,  K.L. (2008): How dying cells alert the immune system to danger. Nat. Rev. Immunol., 8, 279-289.
•  Kääriäinen,  M.,  Järvinen,  T.,  Järvinen,  M.,  Rantanen,  J. and  Kalimo,  H. (2000): Relation between myofibers and connective tissue during muscle injury repair. Scand. J. Med. Sci. Sports, 10, 332-337.
•  Lach,  B. and  Cupler,  E.J. (2008): Macrophagic myofasciitis in children is a localized reaction to vaccination. J. Child Neurol., 23, 614-619.
•  Leeling,  J.L.,  Muni,  I.A.,  Helms,  R.J. and  Johnson,  N. Jr. (1979): Rates of release of subcutaneously injected antigens in the rat. Comparison of an aqueous preparation with two alum-precipated preparations. Allergy, 34, 339-344.
•  Lindblad,  E.B. (2004): Aluminium compounds for use in vaccines. Immunol. Cell Biol., 82, 497-505.
•  Lluís,  F.,  Roma,  J.,  Suelves,  M.,  Parra,  M.,  Aniorte,  G.,  Gallardo,  E.,  Illa,  I.,  Rodríguez,  L.,  Hughes,  S.M.,  Carmeliet,  P.,  Roig,  M. and  Muñoz-Cánoves,  P. (2001): Urokinase-dependent plasminogen activation is required for efficient skeletal muscle regeneration in vivo. Blood, 97, 1703-1711.
•  Lu,  F. and  Hogenesch,  H. (2013): Kinetics of the inflammatory response following intramuscular injection of aluminum adjuvant. Vaccine, 31, 3979-3986.
•  Mann,  C.J.,  Perdiguero,  E.,  Kharraz,  Y.,  Aguilar,  S.,  Pessina,  P.,  Serrano,  A.L. and  Muñoz-Cánoves,  P. (2011): Aberrant repair and fibrosis development in skeletal muscle. Skelet. Muscle, 1, 21.
•  Mannhalter,  J.W.,  Neychev,  H.O.,  Zlabinger,  G.J.,  Ahmad,  R. and  Eibl,  M.M. (1985): Modulation of the human immune response by the non-toxic and non-pyrogenic adjuvant aluminium hydroxide: effect on antigen uptake and antigen presentation. Clin. Exp. Immunol., 61, 143-151.
•  Marichal,  T.,  Ohata,  K.,  Bedoret,  D.,  Mesnil,  C.,  Sabatel,  C.,  Kobiyama,  K.,  Lekeux,  P.,  Coban,  C.,  Akira,  S.,  Ishii,  K.J.,  Bureau,  F. and  Desmet,  C.J. (2011): DNA released from dying host cells mediates aluminum adjuvant activity. Nat. Med., 17, 996-1002.
•  Mark,  A.,  Carlsson,  R.M. and  Granström,  M. (1999): Subcutaneous versus intramuscular injection for booster DT vaccination of adolescents. Vaccine, 17, 2067-2072.
• Maureen, H. and Laura, S. (2012): Anatomy, Physiology, and Behavior, In: The Laboratory Rabbit, Guinea Pig, Hamster, and Other Rodents (Mark, A.S., Karla, A.S., Ronald, P.W. eds), 1st ed., pp. 575-602, Elsevier Academic Press, Amsterdam.
•  Miranda,  S.,  Chaignot,  C.,  Collin,  C.,  Dray-Spira,  R.,  Weill,  A. and  Zureik,  M. (2017): Human papillomavirus vaccination and risk of autoimmune diseases: A large cohort study of over 2million young girls in France. Vaccine, 35, 4761-4768.
•  Mizukami,  T.,  Masumi,  A.,  Momose,  H.,  Kuramitsu,  M.,  Takizawa,  K.,  Naito,  S.,  Maeyama,  J.,  Furuhata,  K.,  Tsuruhara,  M.,  Hamaguchi,  I. and  Yamaguchi,  K. (2009): An improved abnormal toxicity test by using reference vaccine-specific body weight curves and histopathological data for monitoring vaccine quality and safety in Japan. Biologicals, 37, 8-17.
•  Morefield,  G.L.,  Sokolovska,  A.,  Jiang,  D., HogenEsch, H., Robinson, J.P. and Hem, S.L. (2005): Role of aluminum-containing adjuvants in antigen internalization by dendritic cells in vitro. Vaccine, 23, 1588-1595.
•  Mrak,  R.E. (1982): Muscle granulomas following intramuscular injection. Muscle Nerve, 5, 637-639.
•  Murakami,  M. and  Hirano,  T. (2012): The molecular mechanisms of chronic inflammation development. Front. Immunol., 3, 323.
•  Noe,  S.M.,  Green,  M.A., HogenEsch, H. and Hem, S.L. (2010): Mechanism of immunopotentiation by aluminum-containing adjuvants elucidated by the relationship between antigen retention at the inoculation site and the immune response. Vaccine, 28, 3588-3594.
•  Ozawa,  K.,  Hineno,  A.,  Kinoshita,  T.,  Ishihara,  S. and  Ikeda,  S.I. (2017): Suspected Adverse Effects After Human Papillomavirus Vaccination: A Temporal Relationship Between Vaccine Administration and the Appearance of Symptoms in Japan. Drug Saf., 40, 1219-1229.
•  Palmieri,  B.,  Poddighe,  D.,  Vadalà,  M.,  Laurino,  C.,  Carnovale,  C. and  Clementi,  E. (2017): Severe somatoform and dysautonomic syndromes after HPV vaccination: case series and review of literature. Immunol. Res., 65, 106-116.
•  Pittman,  P.R.,  Kim-Ahn,  G.,  Pifat,  D.Y.,  Coonan,  K.,  Gibbs,  P.,  Little,  S.,  Pace-Templeton,  J.G.,  Myers,  R.,  Parker,  G.W. and  Friedlander,  A.M. (2002): Anthrax vaccine: immunogenicity and safety of a dose-reduction, route-change comparison study in humans. Vaccine, 20, 1412-1420.
•  Poland,  G.A.,  Borrud,  A.,  Jacobson,  R.M.,  McDermott,  K.,  Wollan,  P.C.,  Brakke,  D. and  Charboneau,  J.W. (1997): Determination of deltoid fat pad thickness. Implications for needle length in adult immunization. JAMA, 277, 1709-1711.
•  Rabe,  H. (2011): Reference ranges for biochemical parameters in guinea pigs for the Vettest®8008 blood analyzer. Tierarztl. Prax. Ausg. K Klientiere. Heimtiere, 39, 170-175.
•  Shaw,  F.E. Jr.,  Guess,  H.A.,  Roets,  J.M.,  Mohr,  F.E.,  Coleman,  P.J.,  Mandel,  E.J.,  Roehm,  R.R. Jr.,  Talley,  W.S. and  Hadler,  S.C. (1989): Effect of anatomic injection site, age and smoking on the immune response to hepatitis B vaccination. Vaccine, 7, 425-430.
•  Soema,  P.C.,  Kompier,  R.,  Amorij,  J.P. and  Kersten,  G.F. (2015): Current and next generation influenza vaccines: formulation and production strategies. Eur. J. Pharm. Biopharm., 94, 251-263.
• The Vaccine Adverse Reactions Review Committee of Japan. Available from URL:https://www.mhlw.go.jp/stf/shingi/shingi-kousei_284075.html.
•  Tierney,  M.T.,  Aydogdu,  T.,  Sala,  D.,  Malecova,  B.,  Gatto,  S.,  Puri,  P.L.,  Latella,  L. and  Sacco,  A. (2014): STAT3 signaling controls satellite cell expansion and skeletal muscle repair. Nat. Med., 20, 1182-1186.
•  Tomljenovic,  L. and  Shaw,  C.A. (2012): Mechanisms of aluminum adjuvant toxicity and autoimmunity in pediatric populations. Lupus, 21, 223-230.
• U.S. Department of Health and Human Services. The vaccine adverse event reporting system (VAERS). Available from URL:http://vaers.hhs.gov/index
•  Verdier,  F.,  Burnett,  R.,  Michelet-Habchi,  C.,  Moretto,  P.,  Fievet-Groyne,  F. and  Sauzeat,  E. (2005): Aluminium assay and evaluation of the local reaction at several time points after intramuscular administration of aluminium containing vaccines in the Cynomolgus monkey. Vaccine, 23, 1359-1367.
•  Vetrone,  S.A.,  Montecino-Rodriguez,  E.,  Kudryashova,  E.,  Kramerova,  I.,  Hoffman,  E.P.,  Liu,  S.D.,  Miceli,  M.C. and  Spencer,  M.J. (2009): Osteopontin promotes fibrosis in dystrophic mouse muscle by modulating immune cell subsets and intramuscular TGF-beta. J. Clin. Invest., 119, 1583-1594.
•  Weeratna,  R.D.,  McCluskie,  M.J.,  Xu,  Y. and  Davis,  H.L. (2000): CpG DNA induces stronger immune responses with less toxicity than other adjuvants. Vaccine, 18, 1755-1762.
•  Wynn,  T.A. (2008): Cellular and molecular mechanisms of fibrosis. J. Pathol., 214, 199-210.