Immunologic reconstitution following hematopoietic stem cell transplantation despite lymph node paucity in NIK deficiency

Capsule summary:

This case demonstrates successful immune reconstitution following hematopoietic stem cell transplantation in NIK deficiency.

To the Editor.

Antigen-specific adaptive immunity requires interactions between T cells, B cells and dendritic cells in secondary lymphoid organs, which include tonsils, lymph nodes (LNs), Peyer’s patches, and splenic follicles. Activation of the non-canonical NFκB pathway in stromal cells following lymphotoxin β receptor (LTβR) ligation by LTα₁β₂ expressed on lymphoid inducer (LTi) cells is critical for lymphorganogenesis, whereas its activation in B cells after BAFF-R ligation is important for survival¹. LTβR ligation drives NFκB-inducing kinase (NIK) to phosphorylate IKKα¹. IKKα subsequently phosphorylates NF-κB2/p100, which contains an IκB-like domain that prevents nuclear translocation of the p100.RelB complex¹. Phosphorylated p100 undergoes polyubiquitination and proteasomal processing to p52¹. p52:RelB dimers translocate to the nucleus and activate the transcription of genes encoding chemokines and adhesion molecules for lymphorganogenesis, including VCAM1 and CCL20². NIK^−/− and NIK^{Gy860Arg/Gy860Arg} (aly/aly) mice have absent LNs, B cell lymphopenia, disrupted splenic architecture, and hypogammaglobulinemia^{3, 4}. Because p100 is under the transcriptional control of the classical NFκB pathway, activating missense mutations in IκBα disrupt both classical and non-classical NFκB signaling pathways⁵. Consequently, lymphorganogenesis is severely impaired, hematopoietic stem cell transplantation (HSCT) fails to achieve immunologic reconstitution, despite excellent donor cell chimerism, and recurrent infections and need for immunoglobulin replacement persist after HSCT⁵. Two homozygous mutations in NIK, NIK^pro565Arg and NIK^Val344Met, have been described in three patients from two kindreds, who lacked tonsils and palpable LNs^{6, 7}. The only transplanted patient was reported to be well, but no information on immune function was provided⁶. We document robust immunologic reconstitution post-HSCT in a NIK deficient patient with abrogation of non-canonical NFκB signaling, absence of tonsils and paucity of LNs, and we discuss why HSCT may successfully treat NIK deficiency.

Patient 1 (P1), born to consanguineous parents (Figure 1A), presented at one year of age with recurrent sino-pulmonary and gastrointestinal infections and Candida lusitaniae oral infections. He had panhypogammaglobulinemia (IgG <30 mg/dL, IgM <8 mg/dL, IgA <4 mg/dL), undetectable antibody response to tetanus vaccination. He was started on intravenous immunoglobulin replacement (IVIG) and sulfamethoxazole/trimethoprim prophylaxis. At 18 months, he developed a seizure with multiple ring-enhancing lesions in the bilateral cerebral cortexes. Brain biopsy was positive for Mycobacterium with antimicrobial resistance profile characteristic of Bacillus Calmette-Guérin he had been vaccinated with. He was treated with antimycobacterial therapy and prednisone and referred at age 30 months to Boston Children’s Hospital. Physical examination revealed short stature (<3rd percentile), absent tonsils, and no palpable LNs. He had a normal total lymphocyte count, B cell lymphopenia, and severely reduced percentages of CD45RO⁺ memory T cells and IgD^- CD27⁺ switched memory B cells (Table 1). T cell proliferation was normal to PHA, and modestly reduced to tetanus and Candida antigens (Table 1). TLR responses, including TNFα production, were normal (Fig. S1), Serum TNFα was elevated (41 pg/ml). Biopsy of a draining axillary LN was attempted after tetanus re-immunization, but no LNs were identified. To our surprise, lymphoscintigraphy of the lower limb visualized a few right popliteal LNs, but no inguinal LNs (Fig. S2), leading us to pursue HSCT. Nineteen months after antimycobacterial therapy, he demonstrated radiological improvement of brain lesions, and underwent HSCT from an HLA-matched related donor preceded by myeloablative conditioning with busulfan and fludarabine. A year post-HSCT, he is well and demonstrates robust engraftment (>95%) of all hematopoietic lineages. Importantly, he has normal B cell count, normal IgG level, robust pneumococcal and tetanus antibody responses and strong T cell proliferation to tetanus after vaccination. Tonsils and LNs remained undetectable on physical examination. Norovirus in the stools and oral Candida infection cleared post-HSCT. There was no evidence of CMV or EBV infections before, during and after HSCT. P2, P3, and P4 are cousins of P1 (Fig. 1A). They all have recurrent sino-pulmonary infections, bronchiectasis, chronic diarrhea with no pathogens isolated, oral thrush, undetectable tonsils and LNs, and normal lymphocyte and T cell counts. P3 and P4 have B cell lymphopenia. They all have hypogammaglobulinemia and have been placed on IVIG. HSCT has been declined by their parents.

Figure 1. The NIK Cys306Valfs*2 mutation and its impact on signaling.

[A] Family pedigree. [B] Linear map of NIK. The mutation in our patients is indicated in red, and in the two reported kindreds in blue [C] MAP3K14 mRNA expression in fibroblasts from P1 and a healthy control (HC). Data are representative of 3 independent experiments. [D] NIK Protein expression in 293T cells transfected with WT NIK^myc or mutated NIK^myc. [E] p100 processing and phosphorylation in unstimulated and α-LTβR stimulated fibroblasts from P1 and HC. Similar data was obtained in three independent experiments. [F,G] VCAM1 and CCL20 mRNA expression in unstimulated and anti-LTβR stimulated [F] and unstimulated and TNF-α stimulated [G] fibroblasts from P1 and HC. Data are representative of 3 independent experiments. Columns and bars represent the mean ± SEM. *p <.05, **p <.01, and ***p <.001

Table 1.

Immunological profiles of the patients.

Patient age at the time of testing	P1 Pre-HSCT	P1 Post-HSCT	P2	P3	P4
	3.5 year	5.5 year	20 month	5 year	8 year
Lymphocyte subsets
CD3+, 103 cells/μL	4.44 (0.85–4.6)	3.01(0.77–4.0)	3.727 (1.4–7.2)	5.53 (0.85–4.6)	3.77 (0.77–4.0)
CD3+CD4+, 103 cells/μL	3.15 (0.7–2.7)	1.57 (0.4–2.5)	3,359 (0.8–5.2)	4.57 (0.7–2.7)	2.77 (0.4–2.5)
CD45RA+CCR7+, % CD4+	91.3 (65.2–84.8)	56.3 (57.1–84.8)	91.7 (66.3–89)	88.8 (57.1–84.8)	86 (57.1–84.8)
CD45RA+CCR7−, % CD4+	0.1 (0.2–3)	1.3 (0.4–2.6)	1.43 (0.2–2.9)	1.12 (0.4–2.6)	1.43 (0.4–2.6)
CD45RA−CCR7+, % CD4+	6.65 (10.5–23.2)	23.5 (11.2–26.7)	3.79 (1.3–9.4)	3.23 (11.2–26.7)	4.53 (11.2–26.7)
CD45RA−CCR7−, % CD4+	2.1 (2.9–9.8)	18.8 (3.3–15.2)	3.11 (3– 9.4)	6.89 (3.3–15.2)	8.08 (3.3–15.2)
CD3+CD8+, 103 cells/μL	1.23 (0.49–1.3)	1.39 (0.49–1.3)	0.475 (0.2–2.8)	0.99 (0.49–1.3)	0.974 (0.49–1.3)
CD45RA+CCR7+, % CD8+	96.6 (39–89)	14 (28.4–80)	67 (66–89)	84 (28.4–80)	90.4 (28.4–80)
CD45RA+CCR7−, % CD8+	1 (4.8–30)	13.9 (9.1–49.1)	29.3 (6.4–20.8)	12.7 (9.1–49.1)	4.31 (9.1–49.1)
CD45RA−CCR7+, % CD8+	1 (0.9–5.7)	16.6 (1– 4.5)	2.74 (1.7–8.5)	1.49 (1– 4.5)	1.28 (1– 4.5)
CD45RA−CCR7−, % CD8+	1.4 (3.4–28.2)	55.8 (6.2–29.3)	1.02 (5.1–25.1)	0.86 (6.2–29.3)	3.89 (6.2–29.3)
CD19+, 103 cells/μL	0.333 (0.39–1.4)	0.752 (0.39–1.4)	0.61 (0.16–3.7)	0.238 (0.39–1.4)	0.145 (0.39–1.4)
CD27−IgD+, % CD19+	95.11(54–88.4)	96.1 (54–88.4)	91.7 (68–89)	91.5 (54–88.4)	90.1 (47–77)
CD27+IgD+, % CD19+	1.4 (2.7–19.8)	0.8 (2.7–19.8)	1.85 (4.1– 13.9)	0.77 (2.7–19.8)	1.49 (5.2–20.4)
CD27+IgD−, % CD19+	1.8 (3.3–7.4)	2.4 (3.3–7.4)	0.71 (3.9– 13.6)	1.16 (3.3–7.4)	1.49 (10.9–30.4)
CD3−CD56+, 103 cells/μL	0.045 (0.13–0.72)	1.46 (0.13–0.72)	0.222 (0.1–0.8)	0.281 (0.13–0.72)	0.261 (0.13–0.72)
Proliferation (control)
PHA	278,399 (187,403)	123,764 (173,788)	64,091 (45,243)	30,206 (22,965)	20,268 (22,965)
Tetanus toxoid	12,998 (18,084)	100,043 (26,480)	ND	ND	ND
Candida	20,084 (35,821)	209,546 (37,857)	ND	ND	ND
Immunoglobulins (age at time	(1 year)	(5.5 year)	(20 month)	(1 year)	(4 year)
of testing)
IgG, mg/dL	<30	973	<200	<30	228
IgM, mg/dL	10	75	<30	68	32
IgA, mg/dL	<5	16	<5	<5	<5
Tetanus vaccine titer	ND	1.13	ND	ND	ND

ND, not done

UD, undetectable

The values in parentheses for proliferation are those obtained on PBMCS from a normal healthy control studied in parallel and on the same day as patients

Whole exome sequencing of P1, P2 and P3 revealed a homozygous frameshift mutation in MAP3K14 encoding NIK (c.916delT:p.Cys306Valfs*2) within the single region of homozygosity shared by the patients. Sanger sequencing of genomic DNA confirmed the presence of the mutation in all probands (Figure S3). The parents of P1 were heterozygous for the mutation. The frameshift mutation results in substitution of the Cys³⁰⁶ residue with valine, followed by two novel residues and a premature stop codon (Figure 1C). The patient fibroblasts had significantly lower NIK mRNA levels than controls (Figure 1D). Because endogenous NIK protein level is extremely low due to continuous degradation by TRAF3², expression of an N-terminal Myc-tagged mutant NIK was examined in transfected HEK293T cells. Immunoblotting revealed a ~130 kDa band in WT NIK transfected cells and a ~34 kDa band in NIK^{Cys306Valfs*2} transfected cells, corresponding to the expected sizes of WT and mutant NIK (Figure 1E). Fibroblasts from P1 stimulated with α-LTβR for 48 hours had absent p100 phosphorylation and processing (Figure 1F). P1 fibroblasts also failed to upregulate VCAM1 and CCL20 expression after αLTβR stimulation but did so normally after TNF-α stimulation (Figure 1G, H).

LTα can drive the formation of mesenteric and cervical LNs, and splenic PNA+ B cell clusters independently from LTβR signaling in LTβ^−/− mice⁸. Increased TNFα levels can drive mesenteric LNs and organized spleen formation in the absence of LTβR⁹. Intact LTα and TNF-α production and signaling, which depend on the classical NFκB pathway, may have permitted residual lymphoid structures to form in our patient, also reflected in his lack of lymphocytosis and near normal antigen-specific T cell proliferation studies. Residual secondary lymphoid organs (SLOs) may have enabled donor hematopoietic cells to mount a robust adaptive immune response in our patient. As in our patient, HSCT corrects low serum IgG levels in aly/aly mice³. In contrast, in patients and mice with missense mutations in IκBα, disruption of both classical and non-canonical NFκB signaling results in both impaired TNF-α production and defective LTβR signaling, which may not allow the formation of sufficient SLOs for development of an adaptive immune response, resulting in failure of HSCT to achieve immune reconstitution. HSCT may be useful in treating patients with NIK deficiency, or other deficiencies that selectively disrupt the non-canonical NF-κB pathway, who have residual SLOs.

Abbreviations:

CCL20

Chemokine (C-C motif) ligand 20

HSCT

hematopoietic stem cell transplantation

LN

lymph node

NIK

NFκB-inducing kinase

LTα

lymphotoxin α

LTβR

lymphotoxin β receptor

BAFF-R

B-cell activating factor receptor

IKKα

Inhibitor of nuclear factor kappa-B kinase subunit alpha

IκBα

nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha

NFκB

nuclear factor of kappa light polypeptide gene enhancer in B-cells

TRAF3

TNF Receptor Associated Factor 3

VCAM1

Vascular Cell Adhesion Molecule 1