Expression and Clinical Significance of Spi-B in B-cell Acute Lymphoblastic Leukemia

Abstract

Spi-B, a member of the E26 transformation-specific (ETS) family of transcription factors, plays an important role in B cell differentiation. Spi-B also functions in development of diffuse large B-cell lymphoma; thus, we hypothesized that it may participate in leukemogenesis of B-cell acute lymphoblastic leukemia (B-ALL). To test this hypothesis, we first generated an anti-Spi-B monoclonal antibody that recognized Spi-B on formalin-fixed, paraffin-embedded tissue sections. This antibody, designated S28-5, selectively stained B cell nuclei at the pre-plasma cell stage (including centrocytes and centroblasts in germinal centers) and nuclei of plasmacytoid dendritic cells, but not fully differentiated plasma cells, T cells, macrophages, or follicular dendritic cells. Employing S28-5, we then performed immunohistochemical staining of bone marrow aspiration biopsy specimens obtained from B-ALL patients (n=62). Cases that showed stronger nuclear S28-5 signals than T-cell ALL were scored positive. In 26 (42%) of 62 specimens, leukemic cells showed nuclear Spi-B expression, and positivity was associated with patient age at diagnosis, and serum uric acid and creatinine levels. Moreover, Spi-B-positive patients demonstrated significantly shorter overall survival than did Spi-B-negative patients. These results suggest that Spi-B expression may serve as a prognostic indicator of B-ALL.

Introduction

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer. ¹ Approximately 80% of pediatric ALL is of B-cell lineage (B-ALL) and is frequently associated with mutations or chromosomal translocations in target genes encoding transcription factors.^2,3 Although the cure rate in children exceeds 90%, ALL remains an important cause of leukemia-related death, and the prognosis for elderly ALL patients or those with relapsed or refractory ALL remains poor. ⁴ For this reason, pediatric intensified chemotherapy protocols are currently being adapted to adult ALL to improve outcomes, although thus far with limited effect. Therefore, the discovery of new biomarkers to predict or improve prognosis would benefit these ALL patients.

Spi-B (encoded by SPIB) is a nuclear transcription factor of the E26 transformation-specific (ETS) family that plays an important role in B cell differentiation. ³ Spi-B is expressed in B cells at the pre-plasma cell stage and in plasmacytoid dendritic cells, but not in fully differentiated plasma cells, T cells, macrophages, or follicular dendritic cells.^5,6 Like PU.1 (encoded by SPI1), another ETS family transcriptions factor, Spi-B interacts with a purine-rich core sequence, 5’-gAggAA-3.’ ² Human SPIB was cloned from a cDNA library prepared from Raji cells (a human Burkitt lymphoma line) probed with a DNA fragment corresponding to the human PU.1 DNA-binding domain; accordingly, the overall Spi-B amino acid sequence is 43% identical to that of PU.1, and 67% identical in the DNA-binding domain.^5,7

In mice, deletion of the gene encoding Spi-B promotes severe abnormalities in B cell function and T cell-dependent humoral immune responses, and in marked defects in germinal center formation and maintenance. ⁸ Moreover, co-deletion of the gene encoding PU.1 in B-cell lineages results in reduced frequency of B cells in spleen, impaired B cell differentiation, and development of B-ALL with 100% incidence. ⁹ Therefore, Spi-B, acting in concert with PU.1, may function as a tumor suppressor in B cells.

In humans, on the other hand, Spi-B is implicated in lymphomagenesis of diffuse large B-cell lymphoma (DLBCL).^10–12 Lenz et al. ¹¹ reported chromosomal anomalies at the SPIB locus, including gain-of-function mutations/amplifications or translocations, in activated B cell (ABC)-type DLBCL, all of which upregulate SPIB expression. Moreover, they showed that SPIB knockdown by RNA interference was toxic to cell lines derived from ABC-type DLBCL but not to those originating from germinal center B cell (GCB)-type DLBCL. ¹² Furthermore, Takagi et al. ¹³ reported that Spi-B expression, as assessed by immunohistochemistry using the 4G5 anti-human Spi-B monoclonal antibody, correlated with poor prognosis of patients with DLBCL. However, to the best of our knowledge, comparable immunohistochemical analysis has not been performed on human B-ALL, ¹⁴ and it remains unknown whether Spi-B expression can serve as a prognostic biomarker for B-ALL as it does for DLBCL.

In the present study, we asked whether Spi-B is expressed in B-ALL and, if so, whether its expression is associated with B-ALL clinical parameters. To this end, we first generated a monoclonal antibody against human Spi-B that is capable of staining formalin-fixed, paraffin-embedded (FFPE) tissue sections. After rigorous verification of antibody specificity, we conducted immunohistochemical staining of bone marrow aspiration biopsies of B-ALL (n=62) and statistically analyzed potential association between Spi-B expression and clinical parameters and overall survival (OS) of the patients.

Materials and Methods

Construction of Expression Vectors Encoding Spi-B

DNA fragments encoding full-length human Spi-B were polymerase chain reaction (PCR)-amplified with the oligonucleotides 5’-TCAAgCTTCggCACCACCATgCTCgCCCTggA-3’ and 5’-TgCTCgAgTCAggCCCggCGgACTgCAggCAg-3’ (HindIII and XhoI sites underlined) using a human small intestine cDNA library as a template. After HindIII/XhoI digestion, DNA fragments of interest were subcloned into HindIII/XhoI sites of pcDNA3.1/Hygro (Invitrogen, Carlsbad, CA), resulting in pcDNA3.1/Hygro-Spi-B. Using this plasmid as a template, DNA fragments encoding amino acid residues 2-85 of human Spi-B were PCR-amplified with the oligonucleotides 5’-AgggATCCCTCgCCCTggAggCTgCACAgC-3’ and 5’-TAgAATTCTAAgggCTgTAggTggggggTT-3’ (BamHI and EcoRI sites underlined). After BamHI/EcoRI digestion, DNA fragments of interest were subcloned into BamHI/EcoRI sites of pCold-GST (Takara Bio, Kusatsu, Japan), which carries a glutathione S-transferase (GST) tag, resulting in pCold-GST-Spi-B.

Expression and Purification of Spi-B-GST Fusion Protein

Escherichia coli BL21 cells were transformed with pCold-GST-Spi-B and inoculated into 500 mL of Luria-Bertani broth containing 100 µg/mL ampicillin. The cultures were grown at 37C until the optical density at 600 nm reached 0.5. Protein expression was induced by adding isopropyl-β-D-thiogalactopyranoside to a final concentration of 0.3 mM and incubating the culture at 4C overnight. Cells were then harvested from cultures by centrifugation, resuspended in 20 mL phosphate-buffered saline (PBS), and then subjected to freeze-thawing and sonication. After centrifugation at 10,000 x g for 30 min at 4C, supernatants were purified with Glutathione Sepharose 4B (GE Healthcare, Chicago, IL) according to the manufacturer’s instructions. Eluates were again purified using a desalting column PD-10 (GE Healthcare) to eliminate glutathione.

Production of Anti-Spi-B Monoclonal Antibodies

Spi-B-GST fusion protein was mixed with complete Freund’s adjuvant, and emulsion was then injected into the footpad of Wistar Kyoto rats, as described. ¹⁵ Lymphocytes were then collected, mixed with SP2/0 mouse myeloma cells, and electrofused, as described. ¹⁶ Cells were resuspended in regular medium supplemented with 10% BM Condimed H1 (Roche Diagnostics, Mannheim, Germany) and incubated overnight. After adding HAT Media Supplement (Sigma-Aldrich, St. Louis, MO), the cell suspension was dispensed into 96-well plates and cultured until surviving cells formed colonies. Culture media were screened by ELISA using Spi-B-GST fusion protein as the coating antigen, and positive samples were further screened by immunostaining FFPE sections from human terminal ileum tissue. Hybridoma cells in the culture media that selectively stained the nuclei of intestinal microfold cells (M cells) ¹⁷ were cloned by limiting dilution (Fig. 1). Among resulting hybridoma lines, one designated S28-5 was selected for subsequent analysis. The experimental protocol was approved by the Animal Experimentation Committee of Kansai Medical University (reference number 21-094, approved on March 10, 2021).

Figure 1.

Expression of Spi-B in intestinal M cells. Formalin-fixed, paraffin-embedded (FFPE) tissue sections of human terminal ileum were doubly immunostained with S28-5 (red) and GP2-71, an unpublished monoclonal antibody directed against human glycoprotein 2 (GP2), which is an M cell marker (blue). Spi-B expression is detected in nuclei of intestinal M cells expressing cytoplasmic GP2 (arrows). Double immunostaining was performed essentially as described. ¹⁸ Bar=20 µm.

Western Blot Analysis

Human embryonal kidney (HEK) 293T cells, Raji cells and Jurkat cells were obtained from American Type Culture Collection (Manassas, VA). As for 293T cells, 24 hr prior to analysis, cells were transfected with pcDNA3.1/Hygro-Spi-B (or empty pcDNA3.1/Hygro as a control) using the calcium phosphate method. ¹⁹ Cells were lysed by sonication in 1% sodium dodecyl sulfate (SDS) supplemented with protease inhibitors, as described. ²⁰ After incubation at 75C for 10 min, samples were subjected to SDS-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane. After blocking, the membrane was incubated with neat culture medium of S28-5 hybridoma cells at 4C overnight, followed by a 60-min incubation with horseradish peroxidase (HRP)-conjugated anti-rat IgG (Jackson ImmunoResearch, West Grove, PA). The membrane was developed using SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific, Waltham, MA). Subcellular localization of Spi-B was confirmed by immunofluorescence staining, as described. ²¹

Patients and Tissue Samples

Patients were included if they met all of the following criteria: (1) diagnosed with B-ALL at University of Fukui Hospital between 1999 and 2021, (2) availability of FFPE blocks of bone marrow aspiration biopsies collected at the time of first occurrence, and (3) traceable medical records. Consequently, 62 patients were enrolled in this study, and their key clinical information is summarized in Table 1. Instead of obtaining individual written informed consent, an opportunity to opt-out was provided. The study protocol was approved by the Research Ethics Committee of University of Fukui (reference number 20210066, approved on August, 11, 2021).

Table 1.

Association of Spi-B Expression with Clinical Parameters in B-ALL.

	Spi-B-negative	Spi-B-positive	p value
	(n=36)	(n=26)
Age
≥15 years	13 (36.1%)	19 (73.1%)	0.004**
<15 years	23 (63.9%)	7 (26.9%)
Sex
Male	21 (58.3%)	11 (42.3%)	0.213
Female	15 (41.7%)	15 (57.7%)
Leukocytosis a
Present	18 (50.0%)	14 (53.8%)	0.765
Absent	18 (50.0%)	12 (46.2%)
Anemia b
Present	27 (75.0%)	23 (88.5%)	0.186
Absent	9 (25.0%)	3 (11.5%)
Thrombocytopenia c
Present	18 (50.0%)	10 (38.5%)	0.368
Absent	18 (50.0%)	16 (61.5%)
Serum CRP levels
≥1.0 mg/dL	17 (47.2%)	12 (46.2%)	0.128
<1.0 mg/dL	19 (52.8%)	14 (53.8%)
Serum uric acid levels
≥8.0 mg/dL	7 (19.4%)	12 (46.2%)	0.024*
<8.0 mg/dL	29 (80.6%)	14 (53.8%)
Serum creatinine levels
≥0.8 mg/dL	6 (16.7%)	12 (46.2%)	0.012*
<0.8 mg/dL	30 (83.3%)	14 (53.8%)
Serum LDH levels
≥222 U/L	36 (100%)	24 (92.3%)	0.091
<222 U/L	0 (0%)	2 (7.7%)

Abbreviations: CRP, C-reactive protein; LDH, lactate dehydrogenase.

^aWhite blood cell count ≥10,000/µL.

^bHemoglobin ≤11 g/dL.

^cPlatelet count ≤50,000/µL.

^*p<0.05, **p<0.01.

Immunohistochemistry

Immunohistochemical staining for Spi-B was carried out using the Histofine system with modification. ²² Briefly, after deparaffinization, rehydration, and quenching of endogenous peroxidase activity, antigens were retrieved by autoclaving sections in 10 mM citrate buffer (pH 6.0) at 105C for 30 min. After blocking, sections were incubated with S28-5 at 4C overnight. To increase signal intensity, sections were further incubated 30 min with unlabeled mouse anti-rat IgG (Jackson ImmunoResearch) as a “linker” antibody prior to incubation for 60 min with Histofine Simple Stain MAX PO (MULTI) (Nichirei Biosciences, Tokyo, Japan). The color reaction was developed with a Betazoid DAB Chromogen Kit (Biocare Medical, Pacheco, CA).

To detect markers of particular hematopoietic cell populations together with Spi-B in the same tissue sections, we performed double immunohistochemical staining, as described.^18,23 Briefly, sections stained for Spi-B as described above were boiled in 10 mM Tris/HCl buffer (pH 8.0) containing 1 mM EDTA for 3 min to strip bound antibodies and then incubated 60 min with either CD3 (clone PS1; Nichirei Biosciences), CD20 (clone L26; Dako, Glostrup, Denmark), BCL6 (clone LN22; Nichirei Biosciences), CD21 (clone 2G9; Leica Biosystems, Newcastle Upon Tyne, UK), CD68 (clone PG-M1; Dako), CD138 (clone B-A38; Nichirei Biosciences), or CD123 (clone RB4MS; Leica Biosystems), followed by 60-min incubation with Histofine Simple Stain AP (Nichirei Biosciences). The color reaction was developed using a Vulcan Fast Red Chromogen Kit 2 (Biocare Medical). After counterstaining with hematoxylin, sections were mounted with Glycergel mounting medium (Dako).

Statistical Analysis

Associations between Spi-B expression and clinical characteristics of patients were analyzed using the chi-square test. The median OS from the time of initial diagnosis was calculated using the Kaplan-Meier method, and differences between groups were analyzed by the Wilcoxon test and the log-rank test. All analyses were performed using JMP 16 (SAS Institute, Cary, NC) or GraphPad Prism 7 (GraphPad Software, La Jolla, CA). p-values less than 0.05 were considered significant.

Results

The S28-5 Monoclonal Antibody Specifically Recognizes Spi-B

To determine whether S28-5 is specific for Spi-B, we first performed western blot analysis of lysates of 293T cells transfected with pcDNA3.1/Hygro-Spi-B or the empty vector pcDNA3.1/Hygro. As shown in Fig. 2A, a single band of molecular weight ~40 kDa corresponding to Spi-B was detected in cells transfected with cDNA encoding Spi-B (second lane) but absent in mock-transfected cells (first lane). A single band of the same (~40 kDa) molecular weight was detected in B cell lineage Raji cells (Fig. 2A, fourth lane), but was absent in Jurkat cells, which are of the T cell lineage (third lane). Next, to determine subcellular localization of Spi-B recognized by S28-5, we conducted immunofluorescence staining for S28-5 in the same lines. As shown in Fig. 2B (second row), S28-5 selectively stained nuclei of 293T cells transfected with cDNA encoding Spi-B, while staining was absent in mock-transfected cells (top row). S28-5 also stained nuclei of Raji cells (Fig. 2B, bottom row) but not of control Jurkat cells (third row). These results overall indicate that S28-5 selectively recognizes Spi-B expressed in B cell nuclei.

Figure 2.

Specificity of anti-human Spi-B monoclonal antibody S28-5. (A) Western blot analysis of lysates of HEK 293T cells transfected with empty pcDNA3.1/Hygro vector (mock; first lane) or with vector harboring cDNA encoding Spi-B (pcDNA3.1/Hygro-Spi-B) (second lane), or lysates of Jurkat (third lane) or Raji (fourth lane) cells. Membrane was immunoblotted with S28-5. (B) Immunofluorescence staining of HEK 293T cells transfected with pcDNA3.1/Hygro (mock; top row) or with pcDNA3.1/Hygro-Spi-B (second row), or Jurkat (third row) or Raji (bottom row) cells with S28-5. DAPI (4′,6-diamidino-2-phenylindole) was used to mark nuclei (left column). Light blue signals in merged images (right column) indicate nuclear localization of S28-5 signals. Bar=25 µm. (C) Immunofluorescence staining of HEK 293T cells transfected with pcDNA3.1/Hygro (mock; top row) or with pcDNA3.1/Hygro-Spi-B (bottom row) with the commercially available anti-Spi-B monoclonal antibody 4G5. DAPI was used to mark nuclei (left column). Bar=25 µm.

We also assessed specificity of the commercially available anti-Spi-B monoclonal antibody 4G5 (Novus Biologicals, Centennial, CO). To do so, we performed immunofluorescence staining of Spi-B-expressing 293T cells with 4G5 using the same cells and growth conditions used for staining with the S28-5 antibody. As shown in Fig. 2C, unlike S28-5 (see Fig. 2B), 4G5 staining was comparable and cytoplasmic in both Spi-B-expressing and mock-transfected 293T cells, and neither line showed nuclear 4G5 staining.

S28-5 Stains Spi-B-expressing Cell Populations on FFPE Tissue Sections

Montes-Moreno et al. ⁶ previously generated an anti-Spi-B monoclonal antibody 235D and used it to show that Spi-B was exclusively expressed in nuclei of (1) B cells at the pre-plasma cell stage, including centrocytes and centroblasts in germinal centers, and (2) plasmacytoid dendritic cells. To assess whether S28-5 shows a comparable staining pattern, we conducted double immunohistochemical staining for Spi-B and markers of specific hematopoietic cell populations using FFPE sections from normal lymph node tissue. As shown in Fig. 3, S28-5 selectively stained nuclei of CD20+ B cells (panel C) and BCL6+ centroblasts and centrocytes (panel D) but did not stain CD138+ plasma cells (panel E), indicating that Spi-B is preferentially expressed at the pre-plasma cell stage of B cells. Moreover, S28-5 strongly stained nuclei of CD123+ plasmacytoid dendritic cells (Fig. 3, panel F). In contrast, S28-5 did not immunolabel CD3+ T cells (Fig. 3, panel G), CD68+ macrophages (panel H), or CD21+ follicular dendritic cells (panel I). These results are in agreement with those of Montes-Moreno et al. and indicate that S28-5 can be used to detect Spi-B-expressing hematopoietic cells on FFPE tissue sections.

Figure 3.

Spi-B-expressing hematopoietic cell populations in human normal lymph nodes, based on S28-5 staining. (A) Histology of lymphoid follicles in human normal lymph nodes. Hematoxylin and eosin staining. (B) Conventional immunohistochemical staining for Spi-B using the S28-5 antibody. Spi-B-positive cells are localized in and around the germinal center. (C–I) Double immunohistochemical staining for Spi-B and either CD20 (C), BCL6 (D), CD138 (E), CD123 (F), CD3 (G), CD68 (H), or CD21 (I). Spi-B signals were visualized with DAB (brown), and those for the other markers with Fast Red (red). Tissues were counterstained with hematoxylin. Note that nuclei of B cells (inset in C), centroblasts and centrocytes (inset in D) and plasmacytoid dendritic cells (inset in F) are Spi-B-positive, while nuclei of plasma cells (inset in E), T cells (inset in G), macrophages (inset in H) and follicular dendritic cells (inset in I) are Spi-B-negative. Bar=15 µm for insets, and 50 µm for the rest.

Close to Half of All B-ALL Cases Express Spi-B

We then used the S28-5 antibody to conduct immunohistochemical staining of bone marrow aspiration biopsies with B-ALL (n=62). In parallel, we comparably stained samples of T-ALL (n=3) for Spi-B to obtain a signal intensity threshold useful to determine whether B-ALL cases were Spi-B-positive or -negative. As shown in Fig. 4 (top row), blastic cells of T-ALL, which are CD3-positive and CD20-negative, were almost devoid of S28-5 immunolabeling. Thus, we judged CD3-negative/CD20-positive B-ALL specimens with signal intensities equal to or weaker than this as Spi-B-negative (Fig. 4, middle row), and those with signal intensities stronger than this as Spi-B-positive (bottom row). Overall, of 62 B-ALL cases, 26 (42%) were judged Spi-B-positive, and the remaining 36 (58%) were Spi-B-negative.

Figure 4.

Immunohistochemical profile of blastic cells of ALL. Bone marrow aspiration biopsy specimens of T-ALL (top row) and B-ALL (middle and bottom rows) were immunostained for CD3 (left column), CD20 (middle column), or Spi-B (right column), the latter using the S28-5 antibody. Signals were visualized with DAB (brown), and tissues were counterstained with hematoxylin. Representative cases are shown. Bar=10 µm.

As a side note, we also assessed reactivity of the 4G5 antibody against FFPE tissue sections of Spi-B-positive B-ALL. As shown in Fig. 5, while S28-5 stained nuclei of leukemia cells (middle panel), 4G5 only weakly stained the perinuclear cytoplasm and barely stained nuclei (right panel).

Figure 5.

Differential localization of signals detected by two anti-Spi-B monoclonal antibodies, S28-5 and 4G5, in B-ALL leukemia cells. Bone marrow aspiration biopsy specimens of Spi-B-positive B-ALL were stained with hematoxylin and eosin (HE) and immunostained with S28-5 (middle panel) or 4G5 antibody (right panel). Signals were visualized with DAB (brown), and tissues were counterstained with hematoxylin. Bar=10 µm.

Spi-B Expression Is Associated with Age at Diagnosis and Serum Uric Acid and Creatinine Levels in B-ALL

Next, to determine clinical parameters associated with Spi-B expression in B-ALL, we first drew receiver operating characteristic (ROC) curves for all available clinical parameters. We found that (1) age at diagnosis, (2) serum uric acid levels, and (3) serum creatinine levels were associated with Spi-B expression. We then calculated the area under the curve (AUC) of ROC curves for each parameter to determine the optimal cutoff value for each. We found that (1) age of 15 years (AUC of ROC curve 0.746), (2) serum uric acid levels of 8.0 mg/dL (AUC of ROC curve 0.643), and (3) serum creatinine levels of 0.8 mg/dL (AUC of ROC curve 0.682) were optimal cutoff values. Based on these cutoff values, we divided patients into positive and negative groups and conducted chi-square analysis, which revealed the following: (1) the proportion of patients aged ≥15 years in the Spi-B-positive group (73.1%) was greater than that seen in Spi-B-negative group (36.1%) with statistical significance (odds ratio [OR]: 4.80, 95% confidence interval [CI]: 1.596–14.45, p=0.004), (2) the proportion of patients with serum uric acid levels ≥8.0 mg/dL in the Spi-B-positive group (46.2%) was greater than that seen in the Spi-B-negative group (19.4%) with statistical significance (OR: 3.55, 95% CI: 1.148–10.987, p=0.024), and (3) the proportion of patients with serum creatinine levels ≥0.8 mg/dL in the Spi-B-positive group (46.2%) was significantly greater than that seen in the Spi-B-negative group (16.7%) with statistical significance (OR: 4.286, 95% CI: 1.334–13.773, p=0.012) (Table 1).

Spi-B-positivity May Predict a Worse Prognosis for Patients with B-ALL Than Spi-B-negativity

To further assess the relationship between Spi-B expression and patient prognosis in B-ALL, we compared median OS of Spi-B-positive and -negative B-ALL patients. As shown in Fig. 6A, Kaplan-Meier curve analysis combined with the Wilcoxon test revealed that Spi-B-positive patients had significantly shorter median OS than did Spi-B-negative patients (p=0.036), but the difference was not significant when analyzed by the log-rank test (p=0.074). When restricted to adults (≥15 years), Kaplan-Meier analysis showed a similar median OS trend, although differences were not statistically significant (p=0.104 by the Wilcoxon test, and p=0.253 by the log-rank test) (Fig. 6B). By contrast, pediatric patients (<15 years) showed no significant difference in median OS (p=0.279 by the Wilcoxon test, and p=0.277 by the log-rank test), and both Spi-B-positive and -negative patients had a favorable prognosis (Fig. 6C).

Figure 6.

Kaplan-Meier curves showing median overall survival (OS) of patients with B-ALL based on tissue sample Spi-B positivity or negativity, as evaluated by S28-5 staining. Patient groups analyzed included (A) adults (≥15 years; n=32) and children (<15 years; n=30), (B) adults only, and (C) children only.

Discussion

Here, we generated an S28-5 monoclonal antibody that recognizes human Spi-B on FFPE tissue sections. Using it, we performed immunohistochemical analysis of bone marrow aspiration biopsies with B-ALL (n=62) and found that just under half of the cases were Spi-B-positive and that such cases showed a higher age at diagnosis as well as higher serum uric acid and creatinine levels than did Spi-B-negative cases. Furthermore, the median OS of Spi-B-positive patients was significantly shorter than that of Spi-B-negative patients, suggesting that Spi-B expression could be a prognostic biomarker for B-ALL.

Since our initial goal in establishing anti-Spi-B monoclonal antibodies was to stain human intestinal M cells in the follicle-associated epithelium covering Peyer’s patches, as Kanaya et al. ¹⁷ had previously in mice, we screened our antibodies using tissue sections of human terminal ileum (see Materials and methods; see also Fig. 1). However, it was necessary to confirm that S28-5 specifically recognizes Spi-B expressed in hematopoietic cells before we proceeded with staining bone marrow tissues. Therefore, using normal lymph node tissue sections, we confirmed that S28-5 selectively stains B cells at the pre-plasma cell stage (including centrocytes and centroblasts in germinal centers) and plasmacytoid dendritic cells. In all of the above cell types, S28-5 signals were localized almost exclusively in nuclei, with few cytoplasmic signals. On the other hand, fully differentiated plasma cells, T cells, macrophages, and follicular dendritic cells were devoid of S28-5 immunolabeling, as reported previously. ⁶

Takagi et al. ¹³ classified 134 cases of DLBCL as Spi-B-positive (n=26) or -negative (n=108) based on immunohistochemical staining with what was then a commercially available Spi-B antibody (clone 4G5; Abcam, Cambridge, UK) and reported that cases judged Spi-B-positive showed a significantly worse response to treatment and less favorable prognosis than did Spi-B-negative cases. However, positive signals shown in their figure panels appeared localized primarily to the perinuclear cytoplasm (in a Golgi pattern) and not to nuclei. ¹³ In the same study, immunofluorescence analysis using the 4G5 antibody of 293T cells overexpressing full-length FLAG-tagged Spi-B also revealed signals chiefly in the perinuclear cytoplasm, with minimal or no nuclear signals. ¹³ After obtaining the 4G5 antibody, we observed comparable results (see Fig. 2C), but these findings were inconsistent with results seen following S28-5 staining (see Fig. 2B) and also with reports that Spi-B is a nuclear factor.^5,7 Thus, we propose that the 4G5 antibody may recognize antigens other than Spi-B, and therefore its use could lead to erroneous conclusions. Currently, there is no evidence that Spi-B localized in the perinuclear cytoplasm has a function in either normal or lymphoma/leukemia cells.

We found that just under half of B-ALL cases analyzed expressed Spi-B, and to the best of our knowledge, this is the first immunohistological demonstration of Spi-B expression in human B-ALL. Montes-Moreno et al. ⁶ previously proposed that Spi-B can be used as a histological diagnostic marker of blastic plasmacytoid dendritic cell neoplasms. In their report, they also analyzed Spi-B expression in ALL and various histological types of malignant lymphoma and concluded that none of the ALL cases examined (24 cases of B-ALL and 16 cases of T-ALL) were Spi-B-positive. ⁶ Given their rigorous validation of their anti-Spi-B monoclonal antibody 235D, this discrepancy is most likely due to demographic differences in patients analyzed and/or anatomic sites from which tissue samples were derived. Another possibility is that our use of a “linker” antibody (see Materials and methods) for immunohistochemical staining increased assay sensitivity and allowed us to detect nuclear antigens, which are generally difficult to stain.

Relevant to differences between Spi-B-positive and -negative B-ALL cases, it is reported that Spi-B expression levels depend on B cell maturation, namely, that Spi-B expression is lower in pro-B cells and higher in pre-B and mature B cells. ⁷ Thus, leukemic cells in Spi-B-negative cases may be less mature than those in Spi-B-positive cases. However, we did not observe differences in expression levels of B cell surface markers, including CD10, CD19, and CD20, in the two groups (data not shown). Given that Spi-B is a nuclear transcription factor important for B cell differentiation, further studies are needed to assess Spi-B involvement in leukemogenesis in B-ALL.

It is noteworthy that in this study, the proportion of patients with serum uric acid levels ≥8.0 mg/dL or serum creatinine levels ≥0.8 mg/dL in Spi-B-positive group was significantly greater than that seen in Spi-B-negative group. This finding suggests that, compared to Spi-B-negative patients, Spi-B-positive patients may have greater tumor volume and/or renal involvement of B-ALL at the time of diagnosis. It is also possible that elevated serum uric acid and/or creatinine levels are indicators of tumor lysis syndrome. ²⁴ In tumor lysis syndrome, rapid collapse of tumor tissue increases serum uric acid levels, which lead to renal impairment and resulting secondary increases in serum creatinine levels. Although here we observed no significant changes in other biochemical parameters that often fluctuate with development of this syndrome, such as serum potassium, calcium or phosphorus levels (data not shown), it is possible that patients showing Spi-B-positive B-ALL also were affected by tumor lysis syndrome.

Here, we also found that median OS of Spi-B-positive patients was significantly shorter than that of Spi-B-negative patients. We observed a comparable trend in median OS when patients were limited to adults (≥15 years), but those differences were not statistically significant. These results combined suggest that Spi-B expression may be a potential prognostic biomarker for B-ALL. However, considering the fact that pediatric B-ALL patients show favorable prognosis, regardless of Spi-B positivity, this marker may be more useful in predicting prognosis of older B-ALL patients.

Because the Kaplan-Meier curve of the present Spi-B-positive cases showed a sharp drop at early time points, we initially chose the Wilcoxon test, which gives more weight to deaths at early time points. However, the log-rank test, which gives equal weight to all time points, is considered the standard method to test differences between two survival distributions. Therefore, we re-analyzed the same data using the log-rank test and found that differences were not statistically significant (p=0.074). This result suggests that the two groups may exhibit differing responses at early time points and that Spi-B-positive B-ALL cases may be resistant to treatment.

There are some limitations to this study. As this is a single-center retrospective cohort study, our patient number is not large. In addition, since the present cohort was a mix of adults and children, some results obtained may have been influenced by age-related disease characteristics. Future studies focusing on either adult or pediatric disease and including a larger number of patients (which will likely require a multi-center study) may provide more robust results.