This case report highlights the importance of screening for mutations in EPHB4 and other genes that regulate lymphatic development in infants with the nonimmune hydrops fetalis.

Nonimmune hydrops is an important cause of neonatal mortality and in some infants, is associated with developmental anomalies of central lymphatic vessels. The genetic basis of lymphatic developmental defects complicating neonatal hydrops remains undefined. Ephrin type B receptor 4 (EPHB4) is a major regulator of embryonic lymphatic development. To investigate the role of EPHB4 mutation in nonimmune hydrops fetalis, we combined whole-exome sequencing (WES) with histopathology and in vitro functional analysis in an infant with severe neonatal hydrops. Despite active therapy involving thoracentesis and invasive ventilation, the infant eventually succumbed to infection. MRI lymphangiogram revealed complete absence of central lymphatic ducts with dermal lymphatic ectasias. WES identified a rare heterozygous missense mutation in EPHB4 p.Ala700Thr that was maternally inherited. Autopsy revealed pulmonary lymphangiectasia, depletion of lymphoid tissue, and Staphylococcus aureus in spleen. Immunohistochemistry showed loss of prospero homeobox 1 expression in the large lymphatic channels of the lungs and small intestinal villi. In vitro functional studies showed that the EPHB4 mutation resulted in loss of autophosphorylation, decreased extracellular signal-related kinase phosphorylation, and suppressed prospero homeobox 1 expression. This novel case highlights the importance of screening for mutations in EPHB4 and other genes that regulate lymphatic development in infants with the nonimmune hydrops fetalis.

Hydrops fetalis is often a fatal condition, commonly nonimmune in origin (∼85%), and characterized by fluid accumulation in the pleura, pericardium, peritoneum, and skin.1  The etiology of nonimmune hydrops fetalis remains undetermined in 20% of cases.1  In ∼15% of cases of nonimmune hydrops, abnormalities of the lymphatic system, including atresia, lack of central vessels, and defective valve development, have been noted.1,2  Such complex lymphatic anomalies can be categorized into two. The first category includes generalized lymphatic anomaly, Gorham-Stout disease, and Kaposiform lymphangiomatosis that is multifocal with multiorgan involvement.3  The second category is central conducting lymphatic anomaly (CCLA), which comprises central lymphatic dilation and leakage that lead to nonimmune hydrops.3  A genetic basis has been established for some of the conditions.2,4,5  We report a case of fatal nonimmune hydrops with an absence of central lymphatic system due to functional mutation in the ephrin type B receptor 4 (EPHB4) gene. Although associations between EPHB4 mutations and intrauterine deaths, as well as venous/lymphatic dysfunction, in adulthood have been reported, the mechanistic basis of how EPHB4 mutations cause neonatal hydrops fetalis is not fully known.1,2,5 

EPHB4, a venous endothelial cell (EC) surface receptor, is known to play a significant role in embryonic lymphatic development. How mutations in EPHB4 impair development of central lymphatic vessels leading to hydrops remains unclear.2,6  Prospero homeobox 1 (PROX1) is a transcription factor highly expressed in lymphatic ECs, and its loss prevents development of central lymphatics from veins.7,8  We combined whole-exome sequencing (WES), in vivo imaging of lymphatic vessels, postmortem histopathological analysis, and in vitro functional analysis to demonstrate that EPHB4 mutations can cause neonatal hydrops by disrupting PROX1-dependent central lymphatic development.

See the Supplemental Materials for the methods and materials used in this study.

A 35-week gestational age male infant with a birth weight of 4.11 kg was prenatally diagnosed with nonimmune hydrops. The infant was born to a 33-year-old mother with a history of abortion at 11 weeks gestation with cystic hygroma. Mother and grandmother from maternal side had a history of early-onset varicose veins. Aside from varicose veins, mother had swelling of lower extremities only during pregnancy. There was no family history of echocardiography-diagnosed cardiac defects. The infant was hydropic at birth, requiring intubation, and developed recurrent, chylous pleural effusions bilaterally, requiring thoracentesis. The infant was transferred to the level 4 NICU at ∼2 months of age. He had significant subcutaneous edema, and MRI lymphangiogram revealed the absence of central lymphatic vessels and ectasia of subcutaneous lymphatic channels (Fig 1 A and B). A repeat MRI lymphangiogram 60 days later revealed worsening of lymphatic flow to the bowel wall and lumen (Fig 1C). On day of life 145, the infant developed sepsis and succumbed to septic shock, despite intensive cardiovascular and respiratory support.

FIGURE 1

A, Postmortem image showing diffuse subcutaneous edema. B, MRI lymphangiogram showing diffuse ectasia of multiple subcutaneous lymphatic channels over the upper thighs, pelvis, and trunk level of the axilla. Absence of any visible central lymphatic ducts or structure suggests congenital absence or atresia of the central collecting lymphatics. C, Follow-up MRI lymphangiogram 40 days after the previous study, showing increased distension of the subcutaneous, paraspinal, perirectal, and inguinal regions caused by lymphangiectasia. The previously demonstrated lymphatic flow to the bowel wall and lumen was no longer seen.

FIGURE 1

A, Postmortem image showing diffuse subcutaneous edema. B, MRI lymphangiogram showing diffuse ectasia of multiple subcutaneous lymphatic channels over the upper thighs, pelvis, and trunk level of the axilla. Absence of any visible central lymphatic ducts or structure suggests congenital absence or atresia of the central collecting lymphatics. C, Follow-up MRI lymphangiogram 40 days after the previous study, showing increased distension of the subcutaneous, paraspinal, perirectal, and inguinal regions caused by lymphangiectasia. The previously demonstrated lymphatic flow to the bowel wall and lumen was no longer seen.

Close modal

Genetic testing initially started with a karyotype (46XY) and microarray that were noncontributory. WES revealed heterozygous variants in the EPHB4 (locus 7q22.1; GenBank under accession no. NM_004444), HLCS, PKLR, and SOS1 genes. The EPHB4 c.2098G>A (p.Ala700Thr) variant was a missense substitution, with the alanine amino acid being highly conserved in 12 species (up to zebrafish). This maternally inherited EPHB4 variant is predicted to be deleterious/probably damaging, affecting the serine-threonine/tyrosine-protein kinase catalytic domain. HLCS and SOS1 variants were predicted bioinformatically to be tolerant/benign. PKLR variant was intronic and not near a splice site.

Autopsy revealed anasarca, absent thoracic duct, diaphragm microcalcifications, depletion of lymphoid tissue of the reticuloendothelial system, acute global hypoxic-ischemic brain changes with acute meningitis, lungs with extensive bilateral pleural adhesions and pulmonary lymphangiectasia, and postmortem identification of methicillin-resistant Staphylococcus aureus in spleen and Cutibacterium acnes in lung. Immunohistochemistry (IHC) studies showed loss of PROX1 staining in the ECs of the large lymphatic channels in the lung and small intestinal villi, with preservation of podoplanin expression in the same vessels. In comparison, the two postmortem control cases revealed intact PROX1 and podoplanin staining of the lymphatic channels (Figs 2 and 3).

FIGURE 2

IHC of the lung of our proband compared with a 5-month-old matched control patient. Podoplanin in control and case specimens of lung revealed linear endothelial staining of lymphatics, confirming that lymphatic vascular development was normal. Absence of PROX1 staining (a key lymphatic endothelial marker required for lymphatic valve development) in the proband lung showed lack of lymphatic valve development, leading to nonimmune hydrops. Arrows point to ECs.

FIGURE 2

IHC of the lung of our proband compared with a 5-month-old matched control patient. Podoplanin in control and case specimens of lung revealed linear endothelial staining of lymphatics, confirming that lymphatic vascular development was normal. Absence of PROX1 staining (a key lymphatic endothelial marker required for lymphatic valve development) in the proband lung showed lack of lymphatic valve development, leading to nonimmune hydrops. Arrows point to ECs.

Close modal
FIGURE 3

IHC of intestine of our proband compared with a 5-month-old matched control patient. Podoplanin in control and case specimens of intestine revealed linear endothelial staining, confirming that lymphatic vascular development was normal. Absence of PROX1 staining in intestinal specimens of the proband showed lack of lymphatic valve development, leading to nonimmune hydrops. Arrows point to ECs.

FIGURE 3

IHC of intestine of our proband compared with a 5-month-old matched control patient. Podoplanin in control and case specimens of intestine revealed linear endothelial staining, confirming that lymphatic vascular development was normal. Absence of PROX1 staining in intestinal specimens of the proband showed lack of lymphatic valve development, leading to nonimmune hydrops. Arrows point to ECs.

Close modal

The EPHB4 p.Ala700Thr variant localizes to the kinase domain of EPHB47,9  (Fig 4A). We therefore examined EPHB4 phosphorylation in native conditions in HEK293 cells. Although wild-type EPHB4 was significantly phosphorylated (Tyr987) in HEK293 cells, there was an almost complete loss of phosphorylation of the EPHB4 p.Ala700Thr variant (Fig 4B). Similar studies pursued in immortalized human pulmonary microvascular ECs (HPMEC-Im)10  also showed significant attenuation of EPHB4 phosphorylation with the EPHB4 p.Ala700Thr variant (Fig 4 C and D).

FIGURE 4

EPHB4 mutant abolishes signal when transfected into human cells. A, EPHB4 domain map with the mutation shown at amino acid 700 in the kinase domain. B–D, Cells were transfected with wild-type (WT) and mutant (mut) EPHB4 alleles, and lysates were used after 48 hours. B, HEK293 cell culture lysates were used to obtain EPHB4 phosphorylation and EPHB4 by immunoblot (n = 3). C and D, Human lung endothelial cell lysates were used to immunoblot for the phosphorylation of EPHB4 (pEPHB4) and ERK (pERK), as well as EPHB4, ERK, and PROX1, with densitometry shown graphically (n = 4). ***P < .001. CRD, C-type carbohydrate recognition domain; Ctrl, control; FNIII, fibronectin type III; LBP, lipopolysaccharide-binding protein; PMB, pseudomurein cell wall binding; RTK, receptor tyrosine kinase; SAM, sterile alpha motif; TM, transmembrane.

FIGURE 4

EPHB4 mutant abolishes signal when transfected into human cells. A, EPHB4 domain map with the mutation shown at amino acid 700 in the kinase domain. B–D, Cells were transfected with wild-type (WT) and mutant (mut) EPHB4 alleles, and lysates were used after 48 hours. B, HEK293 cell culture lysates were used to obtain EPHB4 phosphorylation and EPHB4 by immunoblot (n = 3). C and D, Human lung endothelial cell lysates were used to immunoblot for the phosphorylation of EPHB4 (pEPHB4) and ERK (pERK), as well as EPHB4, ERK, and PROX1, with densitometry shown graphically (n = 4). ***P < .001. CRD, C-type carbohydrate recognition domain; Ctrl, control; FNIII, fibronectin type III; LBP, lipopolysaccharide-binding protein; PMB, pseudomurein cell wall binding; RTK, receptor tyrosine kinase; SAM, sterile alpha motif; TM, transmembrane.

Close modal

EPHB4 phosphorylation can activate extracellular signal-related kinase (ERK), a mitogen-activated protein kinase important for lymphatic fate specification.7  In HPMEC-Im, although the wild-type allele induced ERK phosphorylation, the EPHB4 p.Ala700Thr variant was not able to further induce ERK phosphorylation (Fig 4 C and D). The transcription factor PROX1 is essentially for maintaining lymphatic EC specification as well as for thoracic and central lymphatic development from venous blood vessels.8,11  We postulated that loss of EPHB4 and ERK phosphorylation would result in decreased PROX1 expression. Immunoblotting studies revealed low PROX1 expression in HPMEC-Im when transfected with empty plasmids (Fig 4 C and D), by strong induction of PROX1 by the EPHB4 wild-type allele, but not by the variant allele (Fig 4 C and D).

Nonimmune hydrops fetalis is frequently associated with major anomalies of the lymphatic system, but the underlying genetic etiology is often not investigated or identified. In this report, we describe a case of fatal nonimmune hydrops with striking radiologic and pathological findings of absence of the central lymphatic system with dermal ectasia of lymphatic vessels consistent with diagnosis of CCLA and associated with a functional mutation in the kinase domain of EPHB4 p.Ala700Thr variant that suppresses self-phosphorylation.12  The EPHB4 mutation is novel because it is not found in the Genome Aggregation Database (https://gnomad.broadinstitute.org). We demonstrate that the variant attenuates ERK phosphorylation and EPHB4-induced expression of PROX1, a transcription factor required for lymphatic EC specification and development of central lymphatic vessels.7,9  Our study reveals a novel role for the EPHB4- PROX1 axis in lymphatic vasculature development that is disrupted by the EPHB4 p.Ala700Thr variant contributing to hydrops fetalis.

After the differentiation of arterial and venous vessel identity, a subset of venous EC acquires the fate of lymphatic EC under the control of transcription factors such as PROX1.1113  Ephrin B2 expressed on arterial ECs works in tandem with its receptor EPHB4 expressed on venous ECs to establish arterial versus venous EC fate.14  EPHB4 plays an important role in maturation of the lymphatic network and lymphatic valve formation in mouse embryos.6  Previously, two EPHB4 mutations, p.Arg739Glu and p.I1e782Ser, were reported to occur in the tyrosine kinase domain of EPHB4 protein, leading to tyrosine kinase inactivity that is required for lymphatic valve development.2  Phenotypes arising from EPHB4 mutations have been reported to be frequently associated with venous malformation, including varicose veins associated with venous stasis and swelling of legs.6  Interestingly, the mother of our proband is a carrier for the EPHB4 mutation with a history of early-onset varicose veins.

In our proband, we demonstrate that the maternally inherited EPHB4 p.Ala700Thr variant was associated with hydrops fetalis. Previously, Li et al5  reported on an infant who developed nonimmune hydrops and was critical in the neonatal period but survived until his mid-20s with a significant history of effusions, ascites, and leg swelling. They performed WES on the proband and a 4-generation pedigree and found an EPHB4 mutation that disrupts a normal splice donor site, resulting in a cryptic splice site that incorporates an additional 4–amino acid sequence into the protein, resulting in loss of EPHB4 phosphorylation as in our proband. It is unclear why the mother did not have symptoms of hydrops in the neonatal period but the proband and the earlier sibling who died in utero had hydrops, as both carried the heterozygous EPHB4 mutation. That EPHB4 mutations can function in an autosomal dominant manner can be an explanation for the phenotype in the proband but not necessarily the milder phenotype in the mother. We speculate that the potential presence of other modifier gene defects or sex-determined effects may have played a role. Martin-Almedina et al2  performed WES in 2 families with a history of in utero neonatal deaths related to nonimmune hydrops and found 2 heterozygous missense variants in EPHB4, localizing to the tyrosine kinase domain, leading to tyrosine kinase inactivity.

Transcriptional programs in lymphatic development are regulated by endothelial ERK signaling.7  The differentiation of venous ECs into lymphatic ECs is regulated by ERK and the transcription factors SOX18, Coup-TFII, and PROX1.7  Our data showed that the EPHB4 variant does not induce further ERK phosphorylation, consistent with repression of ERK signaling. PROX1 is a transcriptional factor required for lymphatic fate specification, and its deletion in mice7,15  prevents ECs from becoming lymphatic ECs, disrupting lymphatic vessel development.7,8,16  Strikingly, EPHB4, but not the protein encoded by the EPHB4 p.Ala700Thr variant, was able to further induce PROX1 expression. To confirm in vivo relevance, IHC studies in our proband and postmortem control cases showed severely reduced PROX1 with EPHB4 mutation.

We describe a novel mutation in EPHB4 that disrupts the gene’s phosphorylation, resulting in decreased PROX1 expression, in a neonate with fatal CCLA. To our knowledge, this study is the first demonstrating a link between EPHB4 mutation and PROX1 expression and CCLA-related hydrops in neonates. Although we report novel data, the mechanisms by which EPHB4 expression regulates PROX1 expression and regulation of lymphatic development by the EPHB4/ERK/PROX1 axis remain to be elucidated. Geneticist referral is warranted because genetic testing algorithms for nonimmune hydrops indicate the performance of next-generation sequencing (whole-genome sequencing or WES) to look for specific genetic mutations in PIEZO1, CCBE1, EPHB4, FOXC2, VEGFR3, SOX18, and ITGA9 genes.17,18  In addition to these genes, WES will inform the discovery of new susceptibility loci for nonimmune hydrops, aiding perinatal management and counseling.

FUNDING: Dr Venkatesh Sampath was supported by Children’s Mercy Research Institute.

Dr Akangire conceptualized and designed the study, collected the clinical data, drafted the initial manuscript, and reviewed and revised the manuscript; Ms Menden and Dr Xia designed the functional studies and interpreted the results and reviewed and revised the manuscript; Dr Thiffault perused the clinical sequencing data from proband and parents and identified the EPHB4 mutation in proband; Dr Ahmed designed the pathology and immunohistochemistry studies and interpreted the results and reviewed and revised the manuscript; Dr Sampath conceptualized and designed the study, supervised data collection, drafted the initial manuscript, and critically reviewed and revised the manuscript for important intellectual content, and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

CCLA

central conducting lymphatic anomaly

EC

endothelial cell

EPHB4

ephrin type B receptor 4

ERK

extracellular signal-related kinase

HEK

human embryonic kidney cell

HPMEC-Im

immortalized human pulmonary microvascular endothelial cells

IHC

immunohistochemistry

PROX1

prospero homeobox 1

WES

whole-exome sequencing

1
Bellini
C
,
Donarini
G
,
Paladini
D
, et al
.
Etiology of non-immune hydrops fetalis: an update
.
Am J Med Genet Part A
.
2015
;
167A
(
5
):
1082
1088
2
Martin-Almedina
S
,
Martinez-Corral
I
,
Holdhus
R
, et al
.
EPHB4 kinase-inactivating mutations cause autosomal dominant lymphatic-related hydrops fetalis
.
J Clin Invest
.
2016
;
126
(
8
):
3080
3088
3
Mäkinen
T
,
Boon
LM
,
Vikkula
M
,
Alitalo
K
.
Lymphatic malformations: genetics, mechanisms and therapeutic strategies
.
Circ Res
.
2021
;
129
(
1
):
136
154
4
Trenor
CC
III
,
Chaudry
G
.
Complex lymphatic anomalies
.
Semin Pediatr Surg
.
2014
;
23
(
4
):
186
190
5
Li
D
,
Wenger
TL
,
Seiler
C
, et al
.
Pathogenic variant in EPHB4 results in central conducting lymphatic anomaly
.
Hum Mol Genet
.
2018
;
27
(
18
):
3233
3245
6
Zhang
G
,
Brady
J
,
Liang
WC
,
Wu
Y
,
Henkemeyer
M
,
Yan
M
.
EphB4 forward signalling regulates lymphatic valve development
.
Nat Commun
.
2015
;
6
:
6625
10.1038/ncomms7625
7
Yu
P
,
Tung
JK
,
Simons
M
.
Lymphatic fate specification: an ERK-controlled transcriptional program
.
Microvasc Res
.
2014
;
96
:
10
15
8
Wigle
JT
,
Oliver
G
.
Prox1 function is required for the development of the murine lymphatic system
.
Cell
.
1999
;
98
(
6
):
769
778
9
Deng
Y
,
Atri
D
,
Eichmann
A
,
Simons
M
.
Endothelial ERK signaling controls lymphatic fate specification
.
J Clin Invest
.
2013
;
123
(
3
):
1202
1215
10
Nitkin
CR
,
Xia
S
,
Menden
H
, et al
.
FOSL1 is a novel mediator of endotoxin/lipopolysaccharide-induced pulmonary angiogenic signaling
.
Sci Rep
.
2020
;
10
(
1
):
13143
11
Jha
SK
,
Rauniyar
K
,
Jeltsch
M
.
Key molecules in lymphatic development, function, and identification
.
Ann Anat
.
2018
;
219
:
25
34
12
Ricci
KW
,
Iacobas
I
.
How we approach the diagnosis and management of complex lymphatic anomalies
.
Pediatr Blood Cancer
.
2021
;
e28985
13
Alitalo
K
,
Tammela
T
,
Petrova
TV
.
Lymphangiogenesis in development and human disease
.
Nature
.
2005
;
438
(
7070
):
946
953
14
Wang
HU
,
Chen
ZF
,
Anderson
DJ
.
Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4
.
Cell
.
1998
;
93
(
5
):
741
753
15
Hong
YK
,
Detmar
M
.
Prox1, master regulator of the lymphatic vasculature phenotype
.
Cell Tissue Res
.
2003
;
314
(
1
):
85
92
16
Srinivasan
RS
,
Oliver
G
.
Prox1 dosage controls the number of lymphatic endothelial cell progenitors and the formation of the lymphovenous valves
.
Genes Dev
.
2011
:
25
(
20
):
2187
2197
17
Michelini
S
,
Paolacci
S
,
Manara
E
, et al
.
Genetic tests in lymphatic vascular malformations and lymphedema
.
J Med Genet
.
2018
;
55
(
4
):
222
232
18
Laterre
M
,
Bernard
P
,
Vikkula
M
,
Sznajer
Y
.
Improved diagnosis in nonimmune hydrops fetalis using a standardized algorithm
.
Prenat Diagn
.
2018
;
38
(
5
):
337
343
10.1002/PD.5243

Competing Interests

CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest to disclose.

Supplementary data