Parvovirus B19 infection and B-cell response pattern in unexplained lymphadenopathy.
Li Gong,Jing Tian,Lanlan Feng,Shumei Wang,Xiaoyan Liu,Gaosheng Huang,Wei Zhang
DOI: https://doi.org/10.1097/CM9.0000000000002789
IF: 6.133
2023-01-01
Chinese Medical Journal
Abstract:Lymphadenopathy is defined as lymph nodes (LN) that are abnormal in size or consistency, usually with a diameter >1 cm. Common etiologies include malignancy, infection, autoimmune disorders, and medications.[1] Most patients obtain a definite diagnosis after undergoing a series of examinations. However, the etiology remains unclear in a small number of patients. Human parvovirus B19 (B19V) is a small, non-enveloped, single-stranded DNA virus, mainly infects erythroid precursor cells in the bone marrow, and causes related diseases. However, it is not known whether B19V causes lymphadenopathy and related clinical problems. Heat shock protein (HSP) 90 is required by many viruses for multiple phases, including viral entry, nuclear import, transcription, and replication, as a highly conserved molecular chaperone,[2] but it is unclear whether HSP90 plays a role in B19V infection. At present, it is generally accepted that B-cell responses are divided into two types of patterns: germinal centers (GCs) and extrafollicular (EF). The former produces long-lived plasma cells and memory B-cells, and the latter produces a rapid response, namely plasmablast reaction.[3] However, it is still unknown for B-cell response in lymphadenopathy with B19V infection. In this study, we observed the clinicopathological characteristics and infection status of B19V in 13 cases of lymphadenopathy using immunohistochemistry, nested polymerase chain reaction (PCR), TaqMan quantitive PCR (qPCR), and electron microscopy. Simultaneously, six cases of lymph nodes without metastatic lung cancer were selected as the control group because it was impossible to obtain normal lymph nodes. The primary antibodies included mouse monoclonal antibodies against B19V VP2 (Clone R92F6, MercMillipore, Temecula, USA), HSP 90 alpha (Clone 2G5.G3, Abcam, Cambridge, UK), Cytomegalovirus (CMV), B-cell lymphoma-2 (Bcl-2) (MX022), CD10 (MX002), CD20 (L26), multiple myeloma oncogene-1 (Mum-1), CD2 (MX074), CD3 (MX036), and rabbit monoclonal antibodies CD5 (SP19) and CD7 (EpB2). All the above antibodies were purchased from MXB Biotechnologies (Fuzhou, China) unless mentioned otherwise. Simultaneously, double immunoenzyme staining with CD20, CD138, and B19V VP2 was performed to further identify cell types, and in situ hybridization (ISH) for Epstein–Barr virus-encoded small RNA (EBER) was used to determine whether EBV infection existed in lymphadenopathy tissues. Analysis of B19V genome was investigated using nested PCR and TaqMan qPCR according to the methods previously described in the literature. In addition, electron microscopy was performed to further provide evidence for B19V infection in lymphadenopathy tissues and one of six lymph nodes as the control group according to the sampling procedures for transmission electron microscopy (Hitachi, HT7800, Japan). Thirteen patients with lymphadenopathy (1.22–2.85 cm in diameter), including seven men and six women, were recruited from Tangdu Hospital, Air Force Medical University, China. Their ages ranged from 15 to 67 years (mean, 35 years). Among them, seven patients presented with fever, and two of which were experiencing skin rash, one patient only presented with skin rash. The other five patients had no symptoms [Supplementary Table 1, https://links.lww.com/CM9/B652]. Histopathologically, all lymph nodes showed reactive hyperplasia with different patterns. The first pattern-GCs pattern, three cases in total, showed follicular hyperplasia of lymphoid follicles. Specifically, hyperplastic lymphoid follicles distributed not only in the cortex but also in the medulla [Figure 1A]. Moreover, the size of some lymphoid follicles was 2–3 times larger than that of normal lymphoid follicles. Immunohistochemically, a few B-cells in GCs were positive for HSP90α positive but were rare outside of GCs [Figure 1B]. However, the B-cells in GCs were widely positive for B19V VP2 [Figure 1C]. In addition, double-labeled immunohistochemical staining showed that the nuclei and membranes of B-cells in GCs were positive for B19V VP2 and CD20 [Figure 1D], respectively.Figure 1: Histopathologically, (A) hyperplastic lymphoid follicles distributed not only in cortex but also in medulla in GCs response pattern. (B) Immunohistochemically, a few B-cells in GCs were positive for HSP90α, but rare outside of GCs. (C)The immunohistochemical staining results of B19V VP2 demonstrated that the nuclei of B-cells in GCs and transformed large cells in the paracortical area were positive. (D) Double-labeled immunohistochemical staining using B19V VP2 and CD20 showed that the nuclei of B-cells from GCs were positive for B19V VP2. Meanwhile, their membranes were positive for CD20. (E) In extrafollicular pattern, lymphoid follicle was hardly seen, and lymphoplasmacytoid cells were dominant. (F) Among them, the transformed cells with abundant cytoplasm and large nuclei were easily found. Moreover, these cells were strongly positive for HSP90α and (G) B19V VP2. (H) Double-labeled immunohistochemical staining showed that the nucleus and membrane of lymphoplasmacytoid cells were positive for B19V VP2 and CD138, respectively. (I–L) The results of electron microscopy showed the presence of B19V particles in the nuclei of DCs and plasmocyte (in four of five cases with B19V DNA positive except that plasmocytes were not found in one case); however, (M, N) B19V particles were only found in DCs from the other five cases of lymphadenopathy. In addition, no virus-like particles were found in lymph tissues as controls (O, P). A, E: Hematoxylin and eosin staining; B–D and F–H: Immunohistochemical staining. A, 63× original magnification; B–H, 400× original magnification; I, K, M, O: 21,000× original magnification; J, L, N, P: 170,000× original magnification. B19V: Human parvovirus B19; CD: Cluster of differentiation; DC: Dentritic cell; GCs: Germinal centers; HSP: Heat shock protein.The second pattern, five cases in total, showed diffuse proliferation and belonged to EF pattern. Lymphoid follicles were rarely found, proliferation of lymphoplasmacytoid cells was dominant, and filled with lymphatic cord. Among them, the transformed cells with abundant cytoplasm and large nucleoli, that is plasmablasts, were readily found [Figure 1E]. Moreover, these cells were strongly positive for HSP90α [Figure 1F] and B19V VP2 [Figure 1G]. Double-labeled immunohistochemical staining showed that the nuclei and cytoplasm of the lymphoplasmacytoid cells in the paracortex were positive for B19V VP2 and CD138 [Figure 1H], respectively. The third pattern, five cases in total, was hyperplasia with lymphoid follicles and the paracortex. The histopathological features and immunohistochemical findings of lymphoid follicles were similar to those of GCs pattern, but the number of hyperplastic follicles decreased as the paracortex was visibly expanded. The immunohistochemical staining of paracortical area showed that there was no loss of T-cell antigen. However, the positive cells for Mum-1 were rich, and transformed cells with abundant cytoplasm and large nuclei were found. In addition, a few large cells with abundant cytoplasm were positive for B19V VP2. The results showed that a low number of plasmablasts were found around the follicles in GCs pattern, but they were frequently observed in the other two patterns. No positive cells were detected for B19V VP2 in six cases of lymph nodes as controls except variable cells from GCs. Similarly, neither EBER nor CMV was positive in all the cases except that only individual cells were positive for EBER in one case. Nested PCR results showed that five of the 13 samples were amplified to the target band of B19V [Supplementary Figure 1, https://links.lww.com/CM9/B652]. TaqMan qPCR amplification demonstrated that the five samples were positive for B19V VP2 genomic region, which further confirmed the above results. Interestingly, all five cases with nested PCR and TaqMan qPCR positivity histologically showed EF pattern. Moreover, four out of five patients presented with fever, and two of them were simultaneously positive for serial EBV capsid/nuclear antigen (EBCA/EBNA) immunoglobin (IgG). However, no amplification was found in six cases of lymph nodes as controls. Except for three cases that were unqualified due to issues with the tissues or slices, the other cases were observed using electron microscopy. The results showed the presence of B19V particles (the range of diameter: 19.3–24.8 nm) in dentritic cells (DCs) and plasmocytes in four out of five cases of lymphadenopathy with B19V DNA positivity except that plasmocytes were not found in one case [Figure 1I–L], confirming B19V infection. However, B19V particles were only found in DCs from the other five cases of lymphadenopathy [Figure 1M and N]. In addition, no virus-like particles were found in the lymph tissues as controls [Figure 1O and P]. It is well known that the receptor and helper receptor of B19V infection only exists in erythroid precursor cells. However, it has been reported that globoside is present in B-cells of GCs.[5] Therefore, B-cells of GCs have a molecular basis for B19V infection. Our recent studies have demonstrated that B19V infection is plausibly an essential contributor to GCs formation in thymus, and the pathogenesis of myasthenia gravis is closely associated with the presence of GCs.[6] However, its danger to humans has not received special attention. In view of histopathology, we referred to GCs responses as follicular hyperplasia pattern and EF responses as parafollicular pattern, and both may cause lymphadenopathy. Moreover, it is generally accepted that GCs-derived plasma cells are the main precursors of long-lived plasma cells to provide long-term protection, whereas EF response is an early, rapid, and strong protective one.[3] Therefore, when the human body is infected by a pathogen, EF response appears first, followed by GCs response. Interestingly, an animal model study has demonstrated that the antiviral immune response is initiated in the lymph nodes with rapid EF production of plasmablasts secreting immunoglobin M, followed by GCs production of IgG-secreting and memory B-cells. In this study, we found that 13 cases of lymphadenopathy displayed GCs, EF, and mixed patterns. Four of five patients with EF pattern had a high fever for a longer time, but no fever was detected in three patients with GCs pattern. Moreover, both B19V DNA and viral proteins were simultaneously detected only in EF pattern, and virus-like particles were found in DCs and plasma cells. These results indicated that B19V infection occurs in this pattern. Furthermore, many plasmacytoid cells and plasmablasts were strongly positive for HSP90α, suggesting that it played a role in B19V replication. However, only B19V protein was detected in the other two patterns. Moreover, virus-like particles were found only in DCs. DCs are also essential for the development of antigen-specific B lymphocytes and plasma cells. Thus, the above results demonstrated that antigens might be transferred to follicular DCs for presentation to B-cells after they were captured by DCs, and further differentiated into plasmablasts and/or plasma cells, forming EF pattern.[3] In addition, we found that plasmablasts were positive for HSP90α in EF. Thus, plasmablasts with HSP90α positivity in EF area play a key role against viruses in the early stages of infection. In the present study, in addition to five cases with EF patterns, we found that B19V protein was positive in the other eight cases of lymphadenopathy, notably in GCs and EF plasmablasts. However, virus-like particles were only found in DCs and not in plasma cells. In addition, only one patient with GCs pattern showed serum B19V IgG antibody positivity. The puzzling question is whether viral DNA may have already been eliminated, as mentioned in our previous study.[5] Considering the above information, the questions are how long EF response lasts before GCs response commences, and whether GCs response appears instantly after EF response. The answer may be negative because the progression of disease has a process. Thus, the third pattern lies between two patterns, and its existence is understandable. In addition, the results of B19V DNA and viral protein detection were similar to GCs pattern, although we noticed that five of the 13 cases belonged to this pattern. This means that it is not in the early stages of infection. Conflicts of interest None.