The activation of BAFF/APRIL system in spleen and lymph nodes of Plasmodium falciparum infected patients

Previous studies have reported activation of the B cell-activating factor (BAFF)/a proliferation-inducing ligand (APRIL) system in T independent immunity against malaria infection. Plasmodium falciparum (P. falciparum) infected animal model is not feasible. Therefore, little is known about the occurrence of BAFF/APRIL system and changes in falciparum lymphoid tissues. This study aimed to investigate the expression of BAFF/APRIL system components in lymphoid tissues from P. falciparum infected patients. Spleen and lymph node samples from 14 patients were collected at autopsy. Normal spleens and bacterially infected tonsils served as controls. The protein and/or mRNA expression of BAFF/APRIL and their cognate receptors, BAFF-R, TACI and BCMA, were determined by immunohistochemistry and RT-qPCR, respectively. The spleens of the patients exhibited significantly higher BAFF-R protein expression than normal spleens. Although without appropriate control, BCMA protein was markedly observed only in the lymph nodes. BAFF and BCMA mRNA levels were also significantly elevated in the spleen tissues of the patients compared with normal spleens. The overall BAFF-R protein levels in the lymphoid tissues of the patients correlated positively with parasitaemia. These findings are the first to confirm that BAFF/APRIL system activation in lymphoid tissues and is positively correlated with the parasitaemia levels in falciparum malaria.

BAFF-R, TACI, and BCMA protein expression in lymphoid tissues from patients with falciparum malaria. The specificity of immunostaining for BAFF-R, TACI, and BCMA was confirmed using appropriate negative and positive controls, including the omission of primary antibodies, and staining of bacterially infected tonsil sections (Fig. 2a) or cultured Raji cells (data not shown), respectively. Representative images of sections stained for BAFF-R, BCMA, and BAFF-R + IgD in sections of normal spleen, spleen and lymph nodes from falciparum malaria patients, and bacterially infected tonsils are shown in Fig. 2a. The intensity of staining is shown in Fig. 2b. Receptor expression levels were evaluated on a semi-quantitative scale (0-400) (Fig. 2c) that incorporated both the intensity of staining and the percentage of cells with positive staining.
Total score of BAFF-R expression was significantly higher in spleen from falciparum malaria patients (200.58 ± 3.74) compared with normal spleen (73.82 ± 10.91). BCMA expression was markedly elevated in the lymph nodes (235.31 ± 5.54), but not the spleens, of falciparum malaria patients ( Fig. 2d and Table 3). Figure 2e shows that the BAFF-R expression scores in the spleens of falciparum malaria patients correlated significantly with parasitaemia (r s = 0.6727, p = 0.039) but not the BCMA expression scores of both lymphoid tissues. In addition, overall BAFF-R protein levels in these lymphoid tissues of patients correlated positively with parasitaemia (r s = 0.5794, p = 0.0117). TACI expression was not detected in sections of the spleen or lymph nodes of falciparum malaria patients, normal spleen, or infected tonsil (data not shown).
Comparing CM and NCM cases, the expression of BAFF-R and BCMA in the spleens or lymph nodes was not significantly different while the parasitaemia level in CM was significantly higher than NCM cases (Supplementary Table 1).
BAff-R, tAci, and BcMA mRnA expression in lymphoid tissues from patients with falciparum malaria. We performed RT-qPCR analysis to measure mRNA levels of the BAFF/APRIL system components in spleen (Fig. 2f), lymph node, and tonsil samples (Fig. 2g). BAFF mRNA expression in the spleens of patients with falciparum malaria was significantly higher than normal spleens (p = 0.0381). In contrast, APRIL mRNA expression tended to downregulate in the spleens of falciparum malaria patients compared with normal spleens (p > 0.05). BAFF-R mRNA expression (fold change) was significantly downregulated in the spleens of falciparum malaria patients (p = 0.0411), whereas BCMA mRNA expression was significantly upregulated only in the spleens of the patients (p = 0.0079) (Fig. 2f). Without comparison with normal relevant tissues, expressions of BAFF, APRIL mRNA and their receptor encoded mRNA in the lymph nodes of the patients were represented as ACTB mRNA ratios 29 (Fig. 2g). The corresponding mRNA expressions in tonsillitis were positive control for RT-qPCR technique (Fig. 2g).

Discussion
Polyclonal B cell activation and alterations in the composition of the B cell compartment are characteristic features of the host response to P. falciparum infection 30,31 . In the present study, we are the first to explore the possibility that the spleen and lymph nodes may be important sites of activation of B cells through the BAFF/APRIL signalling system during P. falciparum infection.
We previously demonstrated the TI production of HZ-specific IgG by human B cells co-cultured with monocytes, which was mediated through activation of the BAFF/APRIL system 32,33 . A number of previous studies have also reported activation of the BAFF/APRIL system by malaria infection; these include studies of patients with P. falciparum malaria 34-37 , patients with P. vivax malaria 36 mice with P. yoelii malaria 38,39 and in vitro studies of B cells stimulated by P. falciparum-derived antigens 33,40 . Muehlenbachs et al. investigated that expression of genes encoding IgG, IgM, CXCL13 (a B cell chemoattractant), and BAFF was increased in Tanzanian placental malaria 34 . Plasma BAFF levels were found to be increased and reflected disease severity in Kenyan children during acute falciparum malaria 35 . In that study, dynamic changes in B cell expression of BAFF-R, BCMA, and TACI were detected during the acute malaria phase and a 16-week follow-up period 35 . An earlier study found that infection of malaria-naïve humans with P. falciparum sporozoites resulted in increased BAFF expression on activated monocytes and dendritic cells, high plasma BAFF levels, and low BAFF-R expression on B cells during the acute phase of the challenge 37 .
The BAFF/APRIL system activation has previously been examined in a study of Brazilian patients infected with P. falciparum and P. vivax 36 . During the acute phase of infection, both species increased plasma APRIL levels whereas only P. falciparum resulted in elevated plasma BAFF levels 36 . The increase in soluble BAFF was in line with reports from other groups 35,37 . TACI expression was increased on T cells but not B cells during the acute phase of P. vivax-induced malaria 36 .
To our knowledge, the present study is the first to examine expression of BAFF/APRIL system components in secondary lymphoid tissue from falciparum malaria patients. We speculate that the elevated BAFF, APRIL, BAFF-R, TACI, and BCMA expressions in the lymphoid tissues reflect activation at the site of antibody production. The determination of these parameters to indicate occurrence of the BAFF/APRIL system activation was consistent with the previous studies in animal malaria 38,39 and other human or animal diseases 3,21,26,27,41,42 and cell cultures 3,43 .
In this study, we detected strong expression of BAFF-R protein in B cells in the spleen and lymph nodes from falciparum malaria patients, consistent with the elevated levels of BAFF mRNA in the same tissues. In contrast, BAFF-R mRNA was significantly downregulated in the tissues of falciparum malaria patients compared with normal spleen, which could be consistent with a high translation rate corresponding to the increase in protein expression. In this study, we could observe TACI mRNA in the positive control tonsillitis and malaria lymph node tissues (Fig. 2g). We did not find significant expression of TACI protein or mRNA in this study (Fig. 2d,f,g) despite the fact that TACI is known to be expressed on mature B cells and plasma cells and binds to both BAFF and APRIL 17 . TACI stimulates differentiation of plasma cells in response to TI-2 antigens 44 . Our failure to examine TACI protein expression was consistent with a previous study that TACI receptor constitutively released from activated B cells which produce soluble decoy receptors 45 .
We detected high expression of BCMA in the B cells of the lymph nodes of falciparum malaria patients, which could be explained by differing transcriptional and/or translational regulation in the two organs. APRIL mRNA, in contrast to BAFF, showed a trend towards downregulation in the spleens of falciparum malaria patients compared with normal spleens, whereas BCMA mRNA expression was significantly upregulated. APRIL is an important signalling molecule for the survival of plasma cells, which express BCMA 46  www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ plasma cells are more dependent on BAFF in the early stages of differentiation, whereas their long-term survival requires APRIL 47,48 . The reason for the observed trend of reduced APRIL mRNA levels is unclear. In our previous study 32 , we observed increased expression of APRIL, BAFF, BAFF-R, TACI, and BCMA in HZ-stimulated monocyte and B cell co-cultures. Therefore, our previous knowledge might contribute that APRIL is also play roles in falciparum malaria.
In addition to its role in the maturation of splenic B cells, BAFF-R is the major mediator of BAFF-dependent co-stimulatory responses in peripheral B and T cells. An earlier study of lymphoid follicles in tonsils revealed an important role for BAFF in the differentiation of germinal centre B cells, and showed that BAFF-R, TACI, and BCMA are differentially modulated during the differentiation of germinal centre B cells to plasma cells 42 .
In this study, we found significant positive correlations between the amount of parasitaemia and BAFF-R protein expression in the lymphoid tissues (spleen and lymph nodes) of falciparum malaria patients (r s = 0.5794, p = 0.0117). Similarly, an earlier study in patients with falciparum placental malaria found that BAFF mRNA was significantly associated with the degree of malarial pigment engulfment by intervillous macrophages 34 .
The present study has some limitations. First, the sample size of spleen and lymph node tissues was small, due in large part to the rarity of autopsies in cases of falciparum malaria. Second, our results reflect the protein/mRNA expression pattern at the time of autopsy, and immunostaining of FFPE autopsied tissue may not accurately represent features of the BAFF/APRIL system in living patients. Third, we lacked normal lymph nodes for comparing changes of each variable expressed in the malaria lymph node tissues. In this study, tonsillitis tissues were only used as positive control for the expression of BAFF/APRIL system and immunostaining of BAFF-R, TACI and BCMA 3 . Fourth, unlike studies with animal models of malaria, we cannot manipulate the BAFF/APRIL system with agonists, antagonists, or blocking specific antibodies in patient tissues.
In conclusion, our results are the first to demonstrate that BAFF/APRIL-mediated B cell activation in the spleen and lymph nodes may play an important role in the immune response to P. falciparum infection. patients and sample collection. Specimens of spleen and/or mesenteric lymph nodes were collected at autopsy from fourteen P. falciparum infected patients who died at The Hospital for Tropical Diseases, Faculty of Tropical Medicine, Mahidol University, between 1973 and 1998. Nine cases died of acute cerebral malaria (CM) whilst in a coma, and five severe cases (non-cerebral malaria, NCM) died of other complications, including pneumonia, septicaemia, acute renal failure, and/or severe anaemia. All CM cases had impaired consciousness as their defining clinical examination criterion for diagnosis of CM according to the WHO guidelines (1991) 49 . In addition, there was a histological finding of parasitized RBCs (PRBCs) in the brain tissues of CM cases. Samples of both spleen and lymph nodes were obtained from five patients, spleen only from five patients, and lymph nodes only from four patients. Normal spleen samples were obtained from five individuals who died from trauma due to traffic accidents between 1996 and 1999. Samples of bacterially infected tonsils were obtained from six individuals. The overall histopathologic changes of the tonsillitis sections stained with haematoxylin & eosin included noticeable congestion and haemorrhage in some areas, and most lymphoid nodules showed activation with an obvious germinal centre. A thickening epithelium was covered with fibrin, necrotic and pus cells, and a number of neutrophils. The clinicopathological characteristics of the fourteen patients are summarised in Table 4.

tissue preparation and histopathological examination. Formalin-fixed, paraffin-embedded (FFPE)
tissues of all samples were sectioned (4 μm thick). Standard histological study with Mayer's haematoxylin and eosin (H&E) was conducted under light microscopy (BX41; Olympus, Japan). P. falciparum-induced histopathological changes in spleen and lymph node sections were categorised based on the severity of gross changes in tissue architecture, alterations to red and white pulp, macrophage infiltration, and PRBC abundance: grade 0, no visible histopathological changes; 1, minimal scattered changes; 2, moderate changes; 3, moderate to severe changes; and 4, severe histopathological changes. As examples, grade 4 changes in a spleen section might include marked congestion and haemorrhage in both white and red pulp, congestive red pulp, severely compressed white pulp, loss of architecture, degenerated central arteries, >80% PRBCs, and >90% macrophages with completely engulfed malarial pigment. In a lymph node section, grade 4 changes might include loss of architecture of the (e) Correlations between parasitaemia and total staining scores for BAFF-R in spleen (left panel) or in both lymphoid tissues (lower panel), or BCMA in spleen of the malaria patients (right panel) using Spearman's rank correlation test with a significance value of p < 0.05. (f) and (g) RT-qPCR analysis of BAFF, APRIL, BAFF-R, TACI, and BCMA mRNA levels in (f) spleens (fold change), and (g) lymph nodes (ACTB mRNA ratio) from patients with falciparum malaria or bacterially infected tonsil (positive control). Data were normalized to ACTB (β-actin) mRNA in the same sample. The results in the spleen of falciparum malaria patients are presented as the fold change in expression relative to the levels in the normal spleen samples using the 2 −(ΔΔCt) method 32 . The results for the lymph node tissues of the patients and tonsillitis were represented as ACTB mRNA ratio 29 . immunohistochemistry. The primary antibodies were rabbit anti-human BAFF-R, goat anti-human TACI, rabbit anti-human BCMA (all from Thermo Fisher Scientific, Waltham, MA, USA), anti-human smooth muscle α-actin (1A4, DAKO, Glostrup, Denmark), and biotinylated anti-human IgD (SouthernBiotech, AL, USA). The secondary antibodies were biotinylated goat anti-IgG (Vector Laboratories, Inc., Burlingame, CA, USA), horseradish peroxidase (HRP)-conjugated rabbit anti-goat IgG (H + L) (Thermo Fisher Scientific). Antibody binding was developed using Vectastain ABC HRP Kit, NovaRed Peroxidase (HRP) Substrate Kit, and DAB Peroxidase (HRP) Kit (all from Vector Laboratories).
After microwave heating antigen-retrieval of tissue sections 50 , immunohistochemistry assay was conducted as described previously 3 and examined by light microscopy (Olympus BX41).
Negative controls with primary antibodies omitted and positive controls (infected tonsil sections and the human B cell line (Raji) were processed in the same manner as above.

Quantification of B cell receptor expression.
For each tissue section, two independent observers (WD and YM) examined a total of 1000 IgD+ B cells in 20 random fields of view under oil immersion. Immunoreactivity for BAFF-R, TACI, and BCMA was evaluated on separate slides. Intensity of staining was scored semi-quantitatively using a 5-point scale: 0, negative; +1, scattered weakly positive; +2, weakly positive; +3, moderately positive; and +4, strongly positive (see Fig. 2b). The total score for each receptor was calculated as the percentage of B cells with positive immunoreactivity multiplied by the intensity score, with minimum and maximum scores of 0 and 400, respectively.
Quantitative reverse-transcription pcR. To determine expression of mRNA encoding BAFF, APRIL, BAFF-R, TACI, BCMA, and β-actin, total RNA was extracted from FFPE tissues as described previously 51,52 . RT-qPCR were performed in duplicate 32,53 using a Luna Universal One-Step RT-qPCR kit (New England Biolabs, MA, USA) and a CFX96TM Real-Time PCR Detection System (Bio-Rad Laboratories, CA, USA). Data were normalized to ACTB (β-actin) mRNA in the same sample. The results in the spleens of the falciparum malaria patients are presented as fold-change in expression relative to the levels in the normal spleen samples using the 2 −(ΔΔCt) method 32 . The results for the lymph nodes of the patients and tonsillitis were represented as ACTB mRNA ratio 29 .  Table 3. Total score of BAFF-R and BCMA expressions in normal spleen, malaria spleen and lymph nodes, and tonsillitis tissues. Note: *Total score of BAFF-R pro tein expression was significant different between malaria and normal spleens using Mann-Whitney U Test. The results were considered statistically significant at the 95% confidence interval (p < 0.05).  Table 2. Histopathological characteristics of lymph node tissue changes in P. falciparum infected patients (9 cases). Note: CM; cerebral malaria, NCM; non-cerebral malaria, ARF; acute renal failure, ARDS; acute respiratory distress syndrome, ND; no data, PRBC; parasitised red blood cells.   49 . In addition, the histological finding of parasitised RBCs (PRBCs) present in the brain tissues of CM cases. NCM; non-cerebral malaria patients; P. falciparum infected patients died of other complications including a pneumonia, b septicaemia, c acute renal failure, and/or d severe anaemia.