Two Cases of Acquired High-Density Lipoprotein Deficiency with Immunoglobulin G4-Related Lecithin–Cholesterol Acyltransferase Autoantibody

Abstract

Lecithin–cholesterol acyltransferase (LCAT) plays a significant role in the progression from premature to mature high-density lipoprotein (HDL) in circulation. Consequently, primary or secondary LCAT deletion or reduction naturally results in low serum HDL cholesterol levels. Recently, rare cases of acquired HDL deficiency with LCAT autoantibodies have been reported, mainly from Japan, where LCAT autoantibodies of immunoglobulin G (IgG) caused the HDL deficiency. Here to our knowledge, we report for the first time two cases of acquired HDL deficiency caused by IgG4 linked LCAT autoantibodies with or without a high serum IgG4 level. Furthermore, these cases can extend to a new concept of “IgG4 autoimmune disease” from the viewpoint of verifying the serum autoantibody and/or renal histopathology.

1. Introduction

It is widely believed that a low concentration of high-density lipoprotein (HDL) particles is an independent risk factor for cardiovascular disease. HDL deficiency is a rare disease with primary or secondary causes. Lecithin–cholesterol acyltransferase (LCAT), via the provision of esterified cholesterol, is a major component in the progression from premature to mature HDL. Thus, a loss or reduction in LCAT function results in abnormal HDL conditions, including low to deficient levels. The three major characteristics of familial LCAT deficiency (FLD) are corneal opacity, anemia, and kidney disease, but there are various conditions of clinical phenotypes by primary gene mutations^{1, 2)}. Fish-eye disease, which is also a primary LCAT gene mutation disease, is characterized by HDL deficiency with severe corneal opacity but without other features. Among variable symptoms, renal failure with nephrotic syndrome that requires dialysis therapy or renal transplantation particularly affects the prognosis in patients with FLD³⁾. There are clear differences in the prognosis between FLD and fish-eye disease, despite the fact that both are primary gene mutation diseases. Conversely, a sudden, severe reduction in HDL by LCAT activity deletion has been rarely reported, in only approximately 10 cases worldwide⁴⁾. Most cases, if not all, were caused by LCAT–immunoglobulin G (IgG) complex, and unlike the primary FLD, an improvement in the natural course or treatments, including prednisolone, can improve the pathogenesis, including kidney disease.

To the best of our knowledge, we report here the first two cases of acquired HDL deficiency with IgG4-related LCAT autoantibody. Furthermore, we demonstrate that these two cases represent the new concept of IgG4-mediated autoimmune disease⁵⁾ that differs from the typical IgG4-related disease with specific organ impairment⁶⁾.

2. Case Presentations

Case 1 was a 70-year-old Japanese man who experienced sudden and extreme reduction in HDL cholesterol (HDL-C) while commuting to the hospital. He had a medical history of percutaneous cardiovascular intervention for angina, diabetes mellitus with insulin treatment 7 years previously, and hypertension, as well as mildly reduced cardiac function. Furthermore, he had no family history of low HDL-C. His serum HDL-C levels were approximately 40 mg/dL but suddenly declined to less than 10 mg/dL and such low HDL-C levels continued. Four months before his HDL-C had started to decrease, his prescriptions had been adjusted so that the amount of metformin was increased from 1250 to 1500 mg for diabetes mellitus treatment, and antidual-platelet therapy with aspirin and clopidogrel was changed to same-component monotherapy 1 month before the drop in HDL-C. The patient’s physical findings and detailed laboratory data were analyzed (Table 1). No symptoms were observed in his physical condition, including corneal opacity, orange-colored and swollen pharyngeal tonsil, or xanthoma. His data revealed extremely low levels of HDL-C and apolipoprotein A-I, which were 3 and 29 mg/dL, respectively. Furthermore, the LCAT activity was not detectable. At this time, the patient was slightly anemic and had slightly elevated serum creatine levels without proteinuria. There were no susceptible data of malignancy, autoimmune disease, or blood disease. Acquired HDL deficiency occurred at age 70 and, together with undetectable LCAT activity, led us to suspect LCAT deficiency caused by autoantibody⁴⁾.

Table 1. Blood profile at consultation in Case 1

			(Reference range)				(Reference range)				(Reference range)
T-Bil	0.7	mg/dL	(0.3-1.2)	ApoA-I	29	mg/dL	(119-155)	ANA	(-)		(-)
AST	20	U/L	(10-40)	ApoA-II	3.5	mg/dL	(25.9-35.7)	RF	16	IU/mL	(≦15)
ALT	12	U/L	(5-45)	ApoB	59	mg/dL	(73-109)	Anti-CCP antibody	0.5	U/mL	(＜4.5)
LDH	160	U/L	(120-245)	ApoC-II	4.0	mg/dL	(1.8-4.6)	CEA	3.9	ng/mL	(≦5.0)
ALP	237	U/L	(104-338)	ApoC-III	8.4	mg/dL	(5.8-10.0)	CA19-9	2.9	U/mL	(≦37.0)
γ-GTP	13	U/L	(≦79)	ApoE	5.4	mg/dL	(2.7-4.3)	SCC	0.9	ng/mL	(≦1.5)
TP	7.4	g/dL	(6.5-8.2)	Lipoprotein fraction				sIL-2R	453	U/mL	(121-613)
Alb	4.2	g/dL	(3.7-5.5)	/agarose gel electrophoresis				ACE	20.6	IU/L	(7.7-29.4)
BUN	17.6	mg/dL	(8.0-20)	α	2	%	(27-51)
Cre	1.15	mg/dL	(0.65-1.09)	preβ＋β	71	%	preβ (8-24)
Na	140	mEq/L	(135-145)				β (35-56)	WBC	4500	/µL	(3500-9700)
K	4.4	mEq/L	(3.5-5.0)	Chylomicron	27	%	(-)	RBC	377	×104/µL	(438-577)
Cl	103	mEq/L	(98-108)					Hb	10.7	g/dL	(13.6-18.3)
UA	6.5	mg/dL	(3.6-7.0)	IgG	1600	mg/dL	(820-1740)	Hct	33	%	(40.4-51.9)
CK	91	U/L	(50-230)	IgG4	136	mg/dL	(11-121)	MCV	88	fL	(83-101)
TC	112	mg/dL	(150-219)	IgA	232	mg/dL	(90-400)	MCH	28	pg	(28.2-34.7)
TG	637	mg/dL	(50-149)	IgM	144	mg/dL	(31-200)	MCHC	32	%	(31.8-36.4)
HDL-C	3	mg/dL	(40-80)					Plt	17	×104/µL	(14.0-37.9)
LCAT	＜70	U	(235-550)
GA	17.4	%	(12.3-16.5)

Abbreviations: T-Bil, total bilirubin; AST, L-aspartate : 2-oxoglutarate aminotransferase; ALT, L-alanine : 2-oxoglutarate aminotransferase; LDH, lactate dehydrogenase; ALP, alkaline phosphatase; γ-GTP, γ-glutamyl transferase; TP, total protein; Alb, albumin, BUN, blood urea nitrogen; Cre, creatinine; Na, sodium; K, potassium; Cl, chlorine; UA, uric acid; CK, creatine kinase; TC, total cholesterol; TG, triglyceride; HDL-C, high density lipoprotein-cholesterol; LCAT, lecithin cholesterol acyltransferase; GA, Glycated albumin; Apo, apolipoprotein; Ig, immunoglobulin; ANA, anti-nuclear antibody; RF, rheumatoid factor; Anti-CCP antibody, anti-cyclic citrullinated peptide antibody; CEA, carcinoembryonic antigen; CA19-9, carbohydrate antigen 19-9; SCC, squamous cell carcinoma-related antigen; sIL-2R, soluble interleukin-2 receptor; ACE, angiotensin-1- converting enzyme WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; Hct, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; Plt, platelet count.

First, the examination for the LCAT autoantibody was checked by enzyme-linked immunosorbent assay (ELISA) analysis. This analysis should have demonstrated almost full inhibition if any LCAT autoantibody was present in the blood as previously described⁷⁾. After the addition of healthy serum or case 1 serum to the ELISA, we observed approximately 50% inhibition of binding to recombinant human LCAT (rhLCAT) on the plate in IgG (Fig.1A, left). Conversely, no inhibition was observed in control (healthy) individuals or in immunoglobulin M (IgM; Fig.1A, right). These outcomes (i.e., not full but partial inhibition to LCAT by IgG) led to the suspicion of the existence of IgG-related LCAT autoantibody, especially derived from the IgG subclass. Next, because the detected LCAT autoantibody had only IgG and not other immunoglobulins, as in previous reports, we investigated the LCAT–IgG complex in the serum using other methods of immunoprecipitation (IP) and Western blotting (WB) as previously described^{4, 8, 9)}. In only the case 1 sample, we detected the LCAT autoantibody using WB (Fig.1B) as compared with the control, after both serum samples were prepared using IP with IgG (IP-IgG). Thus, the IgG-related LCAT autoantibody for low LCAT activity and HDL-C was confirmed in case 1. During follow-up, fibrates were initiated for low HDL-C levels and exacerbation of high triglyceride (TG) levels and proteinuria (Fig.2). However, low HDL-C levels and LCAT activity continued, and the patient’s proteinuria increased to ＞3.5 g/gCr in August 2020 at first visit to nephrology (Fig.2). Therefore, to further confirm the diagnosis, we performed renal biopsy and blood/urine collection (Table 2) in July 2021, which indicated LCAT deficiency disturbance as the cause of proteinuria. At the same time, because heart failure with a large amount of pleural effusion was noted, an investigation of pleural effusion revealed criteria of leaky pleural effusion, and the results provided support for thinking that causes of the pleural effusion were mainly remarkable proteinuria in the nephrotic syndrome range and cardiac dysfunction. After 3 months, steroid therapy with prednisolone was initiated because the patient’s general condition and blood/urine data were continued as above. Subsequently, the HDL-C level quickly and dramatically recovered to the normal range, his remarkable proteinuria was slightly improved after 1 month of steroid treatment (Table 3), and his heart failure improved.

Fig.1. Detection of the LCAT autoantibody in case 1

(A) ELISA analysis. (B) Analysis detected using IP-IgG and WB with or without 2-ME. Abbreviation: H, healthy subject.

Fig.2. Change in lipids and urine profile and overview in case 1

Arrow with gray and No. 1–3 show the point of additional blood data in Supplemental Figure 1.

Table 2. Blood and urine data at renal biopsy in Case 1

			(Reference range)				(Reference range)				(Reference range)
AST	25	U/L	(8-30)	ApoA-I	42	mg/dL	(119-155)	Urine qualitative analysis
ALT	14	U/L	(5-35)	ApoA-II	6.4	mg/dL	(25.9-35.7)	U-protein	4+		(-)
LDH	226	U/L	(100-225)	BNP	554	pg/mL	(≦18.4)	U-glucose	-		(-)
γ-GTP	18	U/L	(7-70)					U-RBC	2+		(-)
TP	6.8	g/dL	(6.5-8.2)	IgG	1301	mg/dL	(870-1700)
Alb	3.7	g/dL	(3.8-5.2)	IgA	243	mg/dL	(110-410)	Urinary sediments
BUN	24	mg/dL	(8.0-20)	IgM	188	mg/dL	(35-220)	RBC	30-49	/HPF
Cre	1.56	mg/dL	(0.61-1.13)	C3	112	mg/dL	(65-135)	Glomerular erythrocyte	+
Na	144	mEq/L	(135-147)	C4	37	mg/dL	(13-35)	WBC	1-4	/HPF
K	4.7	mEq/L	(3.5-5.0)					Granular casts	3-9	/10LPF
Cl	109	mEq/L	(98-108)	WBC	4400	/µL	(3300-8600)	Fatty casts	1-2	/10LPF
CK	110	U/L	(≦160)	RBC	404	×104/µL	(435-555)
TC	129	mg/dL	(130-230)	Hb	10.5	g/dL	(13.7-16.8)	Urine quantitative analysis
TG	497	mg/dL	(30-150)	Hct	32	%	(40.0-50.0)	U-protein＊	7.7	g/gCr	(＜0.15)
HDL-C	15.6	mg/dL	(42-62)	MCV	79	fL	(83.6-98.2)	NAG	86.0	U/L	(≦4.2)
HbA1c	4.3	%	(4.6-6.2)	MCH	26	pg	(27.5-33.2)	β2-microglobulin	2987	ug/L	(≦200)
				MCHC	33	%	(31.7-35.3)
				Plt	20	×104/µL	(15.8-34.8)

Abbreviations: AST, L-aspartate : 2-oxoglutarate aminotransferase; ALT, L-alanine : 2-oxoglutarate aminotransferase; LDH, lactate dehydrogenase; γ-GTP, γ-glutamyl transferase; TP, total protein; Alb, albumin, BUN, blood urea nitrogen; Cre, creatinine; Na, sodium; K, potassium; Cl, chlorine; CK, creatine kinase; TC, total cholesterol; TG, triglyceride; HDL-C, high density lipoprotein-cholesterol; HbA1c, hemoglobin A1c; Apo, apolipoprotein; BNP, B-type natriuretic peptide; Ig, immunoglobulin; C, complement; WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; Hct, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; Plt, platelet count; U-RBC, urine red blood cell; HPF, high power field; LPF, low power field; NAG, N-acethyl-β-D-glucosaminidase. ^＊U-Protein was measured by quantitative analysis (g/gCr).

Table 3. Blood and urine data after 1 month of steroid treatment in Case 1

			(Reference range)				(Reference range)
AST	25	U/L	(8-30)	ApoA-I	126	mg/dL	(119-155)
ALT	17	U/L	(5-35)	ApoA-II	24.1	mg/dL	(25.9-35.7)
LDH	250	U/L	(100-225)
ALP	53	U/L	(38-113)	IgG	915	mg/dL	(870-1700)
γ-GTP	21	U/L	(7-70)	IgG4	77	mg/dL	(4.5-117)
TP	6.3	g/dL	(6.5-8.2)
Alb	3.7	g/dL	(3.8-5.2)
BUN	29	mg/dL	(8.0-20)	WBC	6100	/µL	(3300-8600)
Cre	1.57	mg/dL	(0.61-1.13)	RBC	398	×104/µL	(435-555)
Na	140	mEq/L	(135-147)	Hb	10.0	g/dL	(13.7-16.8)
K	4.4	mEq/L	(3.5-5.0)	Hct	31	%	(40.0-50.0)
Cl	104	mEq/L	(98-108)	MCV	78	fL	(83.6-98.2)
UA	5.7	mg/dL	(2.7-8.0)	MCH	25	pg	(27.5-33.2)
CK	83	U/L	(≦160)	MCHC	32	%	(31.7-35.3)
TC	328	mg/dL	(130-230)	Plt	19	×104/µL	(15.8-34.8)
TG	211	mg/dL	(30-150)
HDL-C	42.5	mg/dL	(42-62)
HbA1c	6.1	%	(4.2-6.2)	U-protein＊	5.4	g/gCr	(＜0.15)

Abbreviations: AST, L-aspartate: 2-oxoglutarate aminotransferase; ALT, L-alanine: 2-oxoglutarate aminotransferase; LDH, lactate dehydrogenase; ALP, alkaline phosphatase; γ-GTP, γ-glutamyl transferase; TP, total protein; Alb, albumin, BUN, blood urea nitrogen; Cre, creatinine; Na, sodium; K, potassium; Cl, chlorine; UA, uric acid; CK, creatine kinase; TC, total cholesterol; TG, triglyceride; HDL-C, high density lipoprotein-cholesterol; HbA1c, hemoglobin A1c;Apo, apolipoprotein; Ig, immunoglobulin; WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; Hct, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; Plt, platelet count; U-protein, urine protein.

^＊U-Protein was measured by quantitative analysis (g/gCr).

Case 2 was a 74-year-old Japanese man with severe low HDL-C levels. The patient had diabetes mellitus, peripheral artery disease, and a history of stroke at 40 years old and underwent coronary artery bypass surgery at 50 years (Table 4). He experienced an episode of sudden HDL-C fall, which is identical to that in case 1. There were no family history of low HDL-C level and no clinical features, such as corneal opacity or tonsil abnormality. Fig.3 demonstrates that the HDL-C level unexpectedly decreased from the normal HDL-C range (40–60 mg/dL) to 1 mg/dL. Five months prior to this episode, his prescription was changed; glucagon-like peptide-1 (GLP-1) agonist was added to his medication along with the discontinuation of one antidiabetic drug due to worsening of diabetes mellitus. The sudden change in the HDL-C level and no family history led us suspect secondary causes. Accordingly, the LCAT autoantibody was also investigated and confirmed using the same methods of WB with IP-IgG sample (Fig.4) as in case 1. During the clinical course, the initiation of fibrate treatment did not provide favorable effects on extremely low HDL-C levels and proteinuria (Fig.3). The general condition of the patient worsened with the progression of peripheral artery disease and occurrence of acute heart failure during follow-up. However, we could not start steroid therapy due to the patient and his family’s unwillingness. Similarly, the patient did not wish to undergo surgery for a severely ischemic limb due to advanced peripheral artery disease and remains under observation.

Table 4. Blood profile in Case 2

			(Reference range)				(Reference range)				(Reference range)
T-Bil	0.5	mg/dL	(0.2-1.2)	LCAT	＜70	U	(235-550)	ANA	160	Folds	(0-40)
AST	29	U/L	(8-38)	Free cholesterol	144	mg/dL	(31-64)	homogeneous pattern	160	Folds	(0-39)
ALT	26	U/L	(4-44)	CE/TC ratio	2	%	(73-77)	Anti-DNA antibody	＜2.0	IU/mL	(0-6)
LDH	234	U/L	(119-229)	ApoA-I	48	mg/dL	(119-155)	Anti-Sm antibody	＜1	Folds	(0-1)
ALP	323	U/L	(104-338)	ApoA-II	6.5	mg/dL	(25.9-35.7)
γ-GTP	24	U/L	(16-73)
TP	6.8	g/dL	(6.7-8.3)					WBC	9520	/µL	(3000-9500)
Alb	3.2	g/dL	(3.9-4.9)	IgG	1722	mg/dL	(870-1700)	RBC	385	×104/µL	(380-550)
BUN	26.5	mg/dL	(8.0-22)	IgG1	900	mg/dL	(351-962)	Hb	11.3	g/dL	(12.0-16.5)
Cre	1.41	mg/dL	(0.6-1.1)	IgG2	829	mg/dL	(239-838)	Hct	35	%	(37-51)
Na	141	mEq/L	(134-147)	IgG3	135	mg/dL	(8.5-140)	MCV	92	fL	(86-104)
K	5.0	mEq/L	(3.6-5.0)	IgG4	59	mg/dL	(4.5-117)	MCH	29	pg	(27.5-36.8)
Cl	108	mEq/L	(98-108)	IgA	331	mg/dL	(110-410)	MCHC	32	%	(31.3-36.6)
UA	5.6	mg/dL	(3.6-8.0)	IgM	59	mg/dL	(33-190)	Plt	23	×104/µL	(13.5-35.5)
CK	98	U/L	(62-287)	C3	92	mg/dL	(65-135)
TC	147	mg/dL	(130-220)	C4	17	mg/dL	(13-35)
TG	296	mg/dL	(30-150)
HDL-C	1	mg/dL	(40-100)
HbA1c	6.5	%	(4.6-6.2)

Abbreviations: T-Bil, total bilirubin; AST, L-aspartate: 2-oxoglutarate aminotransferase; ALT, L-alanine : 2-oxoglutarate aminotransferase; LDH, lactate dehydrogenase; ALP, alkaline phosphatase; γ-GTP, γ-glutamyl transferase; TP, total protein; Alb, albumin; BUN, blood urea nitrogen; Cre, creatinine; Na, sodium; K, potassium; Cl, chlorine; UA, uric acid; CK, creatine kinase; TC, total cholesterol; TG, triglyceride; HDL-C, high density lipoprotein-cholesterol; HbA1c, hemoglobin A1c; LCAT, lecithin cholesterol acyltransferase; CE/TC ratio, cholesteryl ester/total cholesterol ratio; Apo, apolipoprotein; Ig, immunoglobulin; C, complement; ANA, anti-nuclear antibody; WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; Hct, hematocrit; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; MCHC, mean corpuscular hemoglobin concentration; Plt, platelet count.

Fig.3.

Change in lipids and urine profile and overview in case 2

Fig.4. Detection of the LCAT autoantibody in case 2

Analysis detected using IP-IgG and WB with or without 2-ME. Abbreviation: H, healthy subject.

The relationship between IgG4 and autoantibody was recently reported in the field of lipid metabolism¹⁰⁾ and nephrotic syndrome^{11, 12)}. The outcome of our ELISA could indicate the partial association between IgG and LCAT autoantibody in case 1 (Fig.1A, left). Furthermore, high IgG4 serum levels were observed in case 1 (Table 1). Therefore, we focused on the IgG4 in the IgG subclass and proved the existence of IgG4-related LCAT autoantibody in both cases using the identical method of WB with IP-IgG4 sample (Fig.5A), despite the fact that case 2 had a normal IgG4 level (Table 4). Steroid treatment was recently started only in case 1, with HDL-C recovery in the early phase of treatment (Table 3) and serum LCAT protein improvement (Fig.5B), similar to the findings of previous studies^{4, 8)}. However, IgG4-related LCAT autoantibodies still remained 1 month after initiating steroid therapy (Fig.5C).

Fig.5. Detection of the IgG4-related LCAT autoantibody in the two cases and change in serum LCAT protein and autoantibodies between before and after steroid therapy in case 1

(A) Analysis detected using IP-IgG4 and WB with or without 2-ME in cases 1 and 2. (B) Case 1 serum LCAT protein changes between before and after treatment. (C) Case 1 autoantibody analysis before and after therapy, detection using IP-IgG4 and WB with or without 2-ME. Abbreviations: H, healthy subject; Before, before steroid treatment; After, after 1 month of steroid treatment.

In case 1, we performed renal biopsy to determine the cause of large proteinuria, especially with the aim of elucidating the contribution of LCAT deficiency/autoantibody and/or diabetes mellitus. The specimen for light microscopy contained six glomeruli, one of which was sclerosed. Similar to the congenital LCAT deficiency with proteinuria^{2, 13)}, almost all of the glomeruli had mesangial proliferation and lipid deposition and accumulation of foam cells. Furthermore, spike formation, bubbling, and double contouring in the glomerular basement membrane (GBM) were observed (Fig.6A and B). Conversely, compared with the characteristics of FLD, tubular atrophy and interstitial fibrosis were mild. In contrast, there were no obvious histopathological features of diabetic nephropathy, suggesting that diabetes did not contribute to renal insufficiency. Immunofluorescence revealed that IgG (Fig.6C) and IgA were negative; however, there was moderate mesangial IgM (Fig.6D) and C3 deposition. These results did not match with serum IgG4-related LCAT autoantibody in case 1, compared with previous LCAT autoantibody cases^{4, 8)}. Electron microscopy revealed marked lipid droplet accumulation in the GBM and subendothelium (Fig.6E). In addition, we observed thickening of the GBM with subepithelial, intramembranous, and subendothelial electron-dense deposit (Fig.6F). The foot process demonstrated widespread effacement. Those features were consistent with the findings of FLD and acquired LCAT deficiency.

Fig.6. Histological findings of the renal biopsy in case 1

Light microscopy findings of (A) periodic–acid Schiff (PAS) staining and (B) periodic–acid silver–methenamin (PAM) staining. Immunofluorescence findings for (C) IgG and (D) IgM. Electron microscopy revealed (E) abnormal lipid deposits observed in the GBM (closed yellow arrowheads) and (F) GBM thickening with intramembranous electron-dense deposit formation (arrow). Bars: 5 (E) and 2 µm (F).

3.Discussion

The LCAT–IgG complex has been reported among patients with acquired HDL deficiency. However, there have been no reports focusing on the association between the IgG subclass and LCAT autoantibody. In this regard, to the best of our knowledge, we are the first to demonstrate IgG4-related LCAT autoantibody as a cause of secondary HDL deficiency.

IgG4-related disease is unclear in pathogenesis and has the features of high serum IgG4 level with or without antibodies and various organ-specific disturbance with enlargement and nodular hypertrophic lesions⁶⁾. IgG4-related disease is also characterized by infiltration of IgG4-positive plasma cells and lymphocytes to some organs on histopathology. IgG4-related disease is diagnosed using comprehensive criteria in Japan¹⁴⁾, although diagnostic items slightly vary because disorders of individual organs differ clinically. In case 1, computed tomography images and renal biopsy evaluation did not show the presence of localized enlargement, masses, nodules, or thickened lesions, and there was no suspicion of IgG4 invasion into the kidney. Other than a high serum IgG4 level, there were no features that meet the IgG4-related disease criteria in case 1. Similarly, image evaluation indicated no characteristics of IgG4-related disease in case 2. There was only a branch duct intraductal papillary mucinous neoplasm. Moreover, the serum IgG4 level was not increased in case 2. Therefore, IgG4-related disease was not applicable to either case. From the new viewpoint of IgG4 autoimmune diseases, the concept is introduced as represented by the presence of pathogenic, antigen-specific autoantibodies of the IgG4 subclass, for example, those preventing the activity of enzymes or proteins without organ-specific involvement of IgG4 invasion⁵⁾. As one applicable example, IgG4 autoantibody against glycosylphosphatidylinositol-anchored HDL-binding protein 1 (GPIHBP1) in circulation was recently found to be prevalent in patients with high TG^{5, 7)}. Very high TG levels resulting from the inactivation of the GPIHBP1 protein with autoantibody in circulation often induce pancreatitis, as GPIHBP1 is an important factor of TG metabolism. Similarly, this new concept has included the antigen–antibody reaction in circulation, but for some specific organs, it is not necessary for the IgG4 invasion to always be present. In contrast to the detection of IgG4-related antibody for LCAT in serum, there were no suspicious renal tissue findings of the IgG4 complex in case 1 with large proteinuria. However, serum LCAT inhibition by IgG4-related LCAT autoantibody induced renal failure, as in FLD. Thus, case 1 can be included in IgG4 autoimmune diseases. Furthermore, case 2 revealed IgG4-related LCAT autoantibody, even though this patient did not have high circulating IgG4 levels. Case 2 also exhibited the same medical condition of FLD, which matches the new concept. After steroid therapy in case 1, we observed HDL-C recovery (Table 3) and serum LCAT protein improvement (Fig.5B), reflecting release from inhibition of the protein–protein interaction. Surprisingly, IgG4-related LCAT autoantibodies had not disappeared after 1 month of steroid therapy (Fig.5C). To our knowledge, to date, in 2 of 8 acquired LCAT autoantibody cases previously reported⁴⁾, autoantibodies had disappeared after treatment. In one of these two cases, the same method of autoantibody detection, that is, IP-IgG, was used⁴⁾. However, our study used a method specific to IgG4 to detect antibodies, which was more sensitive for IgG4 than for IgG. Therefore, possible reasons for remained autoantibodies could be due to (1) shortness of treatment duration, (2) high sensitivity of the IP method in detecting very small amount of autoantibodies, and (3) in relation to 2), reflecting qualitative, rather than the quantitative nature of the IP method where comparison between before and after should be interpreted with caution, especially at the time of drastic change of both LCAT and IgG4 proteins during steroid therapy. Nevertheless, the observed improvement in LCAT protein/HDL-C levels indicates that steroid therapy reduced autoantibodies enough to increase LCAT function. Some previous reports demonstrated that the appearance of LCAT autoantibody might be associated with specific diseases, for example, non-Hodgkin lymphoma⁹⁾ and sarcoidosis¹⁵⁾. However, cases 1 and 2 did not have obvious disease related to the autoantibody, such as inflammatory, collagen, or hematologic diseases.

Renal failure in patients with FLD has been reported to occur at a median age of 46 years³⁾. The same study reported that the predicting factor for the rapid decline in kidney function is high unesterified cholesterol level. Although secondary LCAT deficiency itself seems to occur in older patients⁴⁾, the progression of renal failure with proteinuria appears to deteriorate in less than 10 years after onset. The strong inactivation of LCAT by autoantibody, whereby inducing high unesterified cholesterol, might result in poor prognosis with rapid progression of renal failure. In addition, renal failure might occur as an effect of both LCAT deficiency and LCAT autoantibody itself in secondary LCAT deficiency. The renal biopsy findings of acquired LCAT deficiency were observed to mimic membranous nephropathy¹³⁾. With regard to the cause of idiopathic membranous nephropathy, IgG4-specific antibody of the phospholipase A2 receptor and the thrombospondin type-1 domain-containing 7A were successively reported^{11, 12)}. Renal histological findings, especially electron microscopy findings of electron-dense deposit for GBM, have indicated that the autoantibody complex might contribute to the causes of proteinuria. However, in case 1, immunofluorescence microscopy revealed a deposit of IgM and C3 but negative for IgG deposition despite IgG4-related LCAT autoantibody in serum. All cases of detected LCAT autoantibody in previous reports had only serum IgG–LCAT autoantibody without other classes of immunoglobulins. Immunofluorescence microscopy of the renal biopsy revealed that some cases were positive for the IgG complex^{4, 8)}, and that the others were negative for all immunoglobulins and complement components^{15, 16)}. Because matched data of the involvement of the LCAT autoantibody in renal failure are not always available, further cases and studies are required to confirm this finding.

The clinical follow-up data of cases 1 and 2 are summarized in Figs.2–3 and Tables 1–4. Fibrates have the actions of not only lowering TG but also elevating HDL. In case 1, the serum HDL-C level was slightly increased and moved by approximately 10–30 mg/dL during the administration of three fibrates (Fig.2). These fluctuations were analyzed in detail via fast protein liquid chromatography (FPLC) to detect lipoprotein fractions (Supplemental Fig.1). The cholesterol of HDL fraction via FPLC was almost not detected when the LCAT activity was below the detection limit (point 1 of Fig.2 and Supplemental Fig.1A). Shortly after the administration of the fibrate, the cholesterol of the HDL fraction was slightly increased when the LCAT activity increased at point 2 of Fig.2 and Supplemental Fig.1B. Despite the continuation of the fibrate, the LCAT activity and cholesterol of the HDL fraction by FPLC decreased again without a change in the serum HDL-C level (point 3 of Fig.2 and Supplemental Fig.1C). Therefore, the fluctuation in the LCAT activity appeared to be approximately more reflected by the cholesterol of the HDL fraction on FPLC than by the serum HDL-C level. In case 1, it was unclear whether the fibrates elevated the serum HDL-C level, as there was a gap between the serum HDL-C and HDL-C fraction on FPLC. In case 2, the serum HDL-C level was not changed at all by fibrate therapy. Thus, the effects of fibrate for lipid metabolism were not evident. Each attending doctor referred to previous studies on urine albumin and lipid profile improvement in FLD¹⁷⁾ and nephrotic syndrome¹⁸⁾, as well as treatment in acquired LCAT deficiency cases^{4, 15)}. In case 1, the attending physician judged that fibrates should be used to treat hypertriglyceridemia (＞500 mg/dL) based on modest increase in serum creatinine levels (1–1.5 mg/dL). He then used all three fibrates (bezafibrate, fenofibrate, pemafibrate). With regard to renal function, neither of the cases reported in this article experienced urine improvement with fibrate treatment.

Supplemental Fig.1. Distribution and alteration of cholesterol and triglyceride levels analyzed via FPLC

The data were (A) at point 1 in Fig. 2, (B) at point 2 in Fig. 2, (C) at point 3 in Fig. 2, and (d) a healthy control.

Abbreviations: VLDL, very-low-density lipoprotein; LDL, low-density lipoprotein; TC, total cholesterol; CE/TC ratio, ratio of cholesteryl ester in total cholesterol.

Because some proteome analyses have indicated that the blood HDL fraction is related to the IgG subclass¹⁹⁾, including IgG4 ²⁰⁾, the IgG4-related autoantibody might be related to HDL itself. However, case 1 demonstrated that the LCAT activity was not associated with the blood IgG4 level (Supplemental Fig.1). Similarly, the IgG4 level itself might not reflect the condition of the LCAT autoantibody, because IgG4 levels remained normal in case 2. Although the main cause of HDL-C fluctuation in case 1 was unclear, it may be plausible to speculate due to the natural course or the effects of the fibrates on autoantibody existence (immunity fluctuation).

Both cases changed prescription before an episode of sudden HDL reduction. In particular, case 2 had started GLP-1 agonist injection, which has an IgG4-crosslinked structure for maintaining a long-acting effect. However, after 6 months, HDL-C was not improved at all with the discontinuation of the GLP-1 agonist. Consequently, the relationship between these drugs and the autoantibody was not clear.

The present study has some limitations. First, the HDL-C value might be different between the laboratories as the direct measurement of HDL-C easily lowers the reliability in severe low HDL cases²¹⁾. Second, because LCAT gene analysis was not performed, we could not completely exclude the possibility of LCAT gene mutation.

In conclusion, to our knowledge, we demonstrated for the first time that two cases of acquired HDL deficiency had IgG4-related LCAT autoantibody regardless of the serum IgG4 level. Furthermore, these pathological conditions may be consistent with the new concept of IgG4 autoimmune disease.

4. Methods

4.1. Blood and Urine Samples and Biochemical Analysis

In both cases, blood and urine were collected and analyzed by each clinical laboratory’s determinations. Free cholesterol and immunoglobulin subclasses were analyzed by the specific laboratories (SRL, Inc., Japan or BML, Inc., Japan). The serum LCAT activity was measured using an exogenous substrate method (Anasorb LCAT, Sekisui Medical, Japan). For lipoprotein fractionation analysis, serum lipoproteins (500 µL total volume) were fractionated using a Superose 6 10/300 GL FPLC column (GE Healthcare Bio-Sciences AB, Sweden) and NGC chromatography system (BIO-RAD Laboratories, Hercules, CA, USA). Fractions were collected and used for lipid measurement. The total cholesterol and TG levels were measured using commercially available assay kits (Wako Pure Chemical Industries, Ltd., Japan).

4.2. Serum LCAT Autoantibody and Protein Detection

First, we used an antibody-based sandwich ELISA method to detect the LCAT autoantibody–immunoglobulin complex in case 1 (at the time of Table 1), as modified in a previous report⁷⁾. Briefly, a 96-well plate was coated with rhLCAT protein, and various diluted serum samples of the control or patient were added. Furthermore, the solution with or without appending a known amount of rhLCAT protein was added and incubated for 1 h. Subsequently, bounded rhLCAT on the plate was detected using horseradish peroxidase (HRP)-labeled antihuman IgG or IgM polyclonal antibody (antihuman IgG-HRP, No. 2040-05; Southern Biotech, Birmingham, AL, USA and antihuman IgM-HRP, No. 2020-05; Southern Biotech). If the serum included the LCAT autoantibody, the LCAT autoantibody was bound to the rhLCAT protein on the plate and in solution. As a result, we determined that the bounded rhLCAT–LCAT autoantibody complex (IgG or IgM) on the plate should be detected and dramatically decreased with inhibition by rhLCAT in solution.

Next, the IgG or IgG4-related LCAT autoantibody was detected in case 1 (at the time of Table 1) and case 2 (at the time of Table 4) using the IP and WB methods, as described in a previous report⁴⁾. When performing the IP, the beads were used for IgG (Dynabeads Protein G; Thermo Fisher Scientific, Waltham, MA, USA) and IgG4 (Dynabeads M-280 sheep antimouse IgG; Thermo Fisher Scientific and mouse antihuman IgG4 Fc monoclonal antibody MAB1312; Merck, Germany). The beads, antibody, and each serum-matched total protein level via BCA protein assay (Thermo Fisher Scientific, Rockford, IL, USA) or matched serum volumes were mixed in accordance with the protocols. After IP was performed for each IgG and IgG4, each sample, prepared with or without 2-mercaptoethanol (2-ME), was then subjected to Western blot analysis using the devices (Bio-Rad Laboratories). For autoantibody detection, electrophoresis was also performed, under nonreducing conditions, suggesting that 2-ME was not included, because some autoantibodies are known to disappear due to the reducing action¹¹⁾. For detecting serum LCAT protein, Western blot analysis was performed using 100-fold diluted serum with general reducing condition. The LCAT antibody (ab51060, Abcam, UK; 1:10,000, overnight) and antirabbit IgG-HRP conjugate (PI-1000; Vector Laboratories, Inc., Newark, CA, USA; 1:5000, 1 h) were used for detection after loading per lane, separation by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, transferring the membranes, and blocking.

Acknowledgements

This work was supported in part by funding from Fukuoka University (Grant No. 207108).

Conflict of Interest

Dr. Uehara received research funding from Asahi Kasei Co. LTD. and Mizuno Co. LTD.

Other authors have no conflicts of interest to disclose.

References

• 1)  Ng DS: The role of lecithin: cholesterol acyltransferase in the modulation of cardiometabolic risks - a clinical update and emerging insights from animal models. Biochim Biophys Acta, 2012; 1821: 654-659
• 2)  Kuroda M, Bujo H, Yokote K, Murano T, Yamaguchi T, Ogura M, Ikewaki K, Koseki M, Takeuchi Y, Nakatsuka A, Hori M, Matsuki K, Miida T, Yokoyama S, Wada J and Harada-Shiba M: Current Status of Familial LCAT Deficiency in Japan. J Atheroscler Thromb, 2021; 28: 679-691
• 3)  Pavanello C, Ossoli A, Arca M, D’Erasmo L, Boscutti G, Gesualdo L, Lucchi T, Sampietro T, Veglia F and Calabresi L: Progression of chronic kidney disease in familial LCAT deficiency: a follow-up of the Italian cohort. J Lipid Res, 2020; 61: 1784-1788
• 4)  Ishibashi R, Takemoto M, Tsurutani Y, Kuroda M, Ogawa M, Wakabayashi H, Uesugi N, Nagata M, Imai N, Hattori A, Sakamoto K, Kitamoto T, Maezawa Y, Narita I, Hiroi S, Furuta A, Miida T and Yokote K: Immune-mediated acquired lecithin-cholesterol acyltransferase deficiency: A case report and literature review. J Clin Lipidol, 2018; 12: 888-897
• 5)  Koneczny I: Update on IgG4-mediated autoimmune diseases: New insights and new family members. Autoimmun Rev, 2020; 19: 102646
• 6)  Umehara H, Okazaki K, Kawa S, Takahashi H, Goto H, Matsui S, Ishizaka N, Akamizu T, Sato Y, Kawano M, Research Program for Intractable Disease by the Ministry of Health L and Welfare J: The 2020 revised comprehensive diagnostic (RCD) criteria for IgG4-RD. Mod Rheumatol, 2021; 31: 529-533
• 7)  Beigneux AP, Miyashita K, Ploug M, Blom DJ, Ai M, Linton MF, Khovidhunkit W, Dufour R, Garg A, McMahon MA, Pullinger CR, Sandoval NP, Hu X, Allan CM, Larsson M, Machida T, Murakami M, Reue K, Tontonoz P, Goldberg IJ, Moulin P, Charriere S, Fong LG, Nakajima K and Young SG: Autoantibodies against GPIHBP1 as a Cause of Hypertriglyceridemia. N Engl J Med, 2017; 376: 1647-1658
• 8)  Takahashi S, Hiromura K, Tsukida M, Ohishi Y, Hamatani H, Sakurai N, Sakairi T, Ikeuchi H, Kaneko Y, Maeshima A, Kuroiwa T, Yokoo H, Aoki T, Nagata M and Nojima Y: Nephrotic syndrome caused by immune-mediated acquired LCAT deficiency. J Am Soc Nephrol, 2013; 24: 1305-1312
• 9)  Simonelli S, Gianazza E, Mombelli G, Bondioli A, Ferraro G, Penco S, Sirtori CR, Franceschini G and Calabresi L: Severe high-density lipoprotein deficiency associated with autoantibodies against lecithin: cholesterol acyltransferase in non-Hodgkin lymphoma. Arch Intern Med, 2012; 172: 179-181
• 10)  Miyashita K, Lutz J, Hudgins LC, Toib D, Ashraf AP, Song W, Murakami M, Nakajima K, Ploug M, Fong LG, Young SG and Beigneux AP: Chylomicronemia from GPIHBP1 autoantibodies. J Lipid Res, 2020; 61: 1365-1376
• 11)  Beck LH, Jr., Bonegio RG, Lambeau G, Beck DM, Powell DW, Cummins TD, Klein JB and Salant DJ: M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy. N Engl J Med, 2009; 361: 11-21
• 12)  Tomas NM, Beck LH, Jr., Meyer-Schwesinger C, Seitz-Polski B, Ma H, Zahner G, Dolla G, Hoxha E, Helmchen U, Dabert-Gay AS, Debayle D, Merchant M, Klein J, Salant DJ, Stahl RAK and Lambeau G: Thrombospondin type-1 domain-containing 7A in idiopathic membranous nephropathy. N Engl J Med, 2014; 371: 2277-2287
• 13)  Najafian B, Lusco MA, Finn LS, Alpers CE and Fogo AB: AJKD Atlas of Renal Pathology: Lecithin-Cholesterol Acyltransferase (LCAT) Deficiency. Am J Kidney Dis, 2017; 70: e5-e6
• 14)  IgG4-related diseases on the Japan Intractable Diseases Information Center’s website. https://www.nanbyou.or.jp/entry/4504 (in Japanese)
• 15)  Akiko T, Okura T, Nagao T, Kukida M, Enomoto D, Miyoshi KI, Higaki J, Kuroda M and Bujo H: A case of acquired lecithin: cholesterol acyltransferase deficiency with sarcoidosis that remitted spontaneously. CEN Case Rep, 2016; 5: 192-196
• 16)  Shoji K, Morita H, Ishigaki Y, Rivard CJ, Takayasu M, Nakayama K, Nakayama T, Inoue Y, Ayaki M and Yoshimura A: Lecithin-cholesterol acyltransferase (LCAT) deficiency without mutations in the coding sequence: a case report and literature review. Clin Nephrol, 2011; 76: 323-328
• 17)  Yee MS, Pavitt DV, Richmond W, Cook HT, McLean AG, Valabhji J and Elkeles RS: Changes in lipoprotein profile and urinary albumin excretion in familial LCAT deficiency with lipid lowering therapy. Atherosclerosis, 2009; 205: 528-532
• 18)  Agrawal S, Zaritsky JJ, Fornoni A and Smoyer WE: Dyslipidaemia in nephrotic syndrome: mechanisms and treatment. Nat Rev Nephrol, 2018; 14: 57-70
• 19)  Zheng JJ, Agus JK, Hong BV, Tang X, Rhodes CH, Houts HE, Zhu C, Kang JW, Wong M, Xie Y, Lebrilla CB, Mallick E, Witwer KW and Zivkovic AM: Isolation of HDL by sequential flotation ultracentrifugation followed by size exclusion chromatography reveals size-based enrichment of HDL-associated proteins. Sci Rep, 2021; 11: 16086
• 20)  Davidson WS, Shah AS, Sexmith H and Gordon SM: The HDL Proteome Watch: Compilation of studies leads to new insights on HDL function. Biochim Biophys Acta Mol Cell Biol Lipids, 2022; 1867: 159072
• 21)  Miida T, Nishimura K, Okamura T, Hirayama S, Ohmura H, Yoshida H, Miyashita Y, Ai M, Tanaka A, Sumino H, Murakami M, Inoue I, Kayamori Y, Nakamura M, Nobori T, Miyazawa Y, Teramoto T and Yokoyama S: Validation of homogeneous assays for HDL-cholesterol using fresh samples from healthy and diseased subjects. Atherosclerosis, 2014; 233: 253-259