Introduction

Isolated methionine synthase (MS) deficiency is an inborn error of intracellular cobalamin metabolism. The MTR gene encodes MS and is localised on chromosome 1q43 [7]. MS catalyses the transmethylation of homocysteine to methionine using cobalamin as a cofactor and 5-methyltetrahydrofolate to form methylcobalamin (Fig. 1a). Somatic cell complementation analysis showed that MS deficiency can be divided into two complementation groups, cblE and cblG [13].

Fig. 1
figure 1

a Methionine synthase reaction and link to folate metabolism. Methionine synthase catalyses the conversion of homocysteine to methionine and needs methylcobalamin as a cofactor. The steps affected in the cblG and cblE defects as well as MTHFR deficiency are shown. AdoMet S-Adenosylmethionine, THF tetrahydrofolate, MTHFR methylentetrahydrofolate reductase. b Fundus photographs taken with a RetCam system (Massie Laboratories, Pleasanton, CA, USA). In both eyes, the macula appears to be normal. Both optic discs have sharp margins but a pale appearance. OD Right eye, OS left eye

Full size image

The disease manifests in the first years of life with neurological symptoms, hyperhomocysteinaemia, homocystinuria, low levels of methionine and megaloblastic anaemia. Visual symptoms include nystagmus and blindness [1].

To our best knowledge only 28 patients with the cblG defect have been reported so far [11, 12]. In this report we describe a new case with special emphasis on the impact of functional MS deficiency on the visual system.

Case report

Material and methods

A complete ophthalmologic examination was performed. Ganzfeld electroretinograms (ERG) and flash visual evoked potentials (VEP) were recorded according to the current standards of the International Society for Clinical Electrophysiology of Vision (ISCEV) using a Nicolet system (Nicolet Biomedical, Madison, WI, USA). Time-integrated luminance of the standard flash (SF) was 3.0 photopic cd s/m2. For derivation of the rod response it was 2.5 log units below the SF. Ganzfeld ERGs were recorded at approximately yearly intervals, at age 14 months, 2.5 and 3.5 years. Due to scheduling difficulties the last examination was performed at age 5.5 years. Flash VEPs were recorded at age 14 months and 3.5 years. Biomicroscopy, retinoscopy and ERGs were performed under anaesthesia using sevoflurane due to low compliance.

Complementation analysis [13] confirmed the cblG defect revealed by methionine and serine formation studies [3, 4].

Patient

The male child was born at 36 weeks of gestation. Magnetic resonance imaging (at age 2 months) showed a combined internal and external communicating hydrocephalus. Eight weeks later obstructive hydrocephalus developed, requiring a ventriculo-peritoneal shunt. Postoperatively, pneumonia developed that resolved under intravenous therapy with cefotiam. Severe global developmental delay became apparent and homocystinuria was present. Plasma levels of total homocysteine were elevated at 73 μmol/l. Under substitution therapy homocystine became undetectable in urine. However, plasma homocysteine remained elevated at 90 μM (age 7 months). An EEG (age 10 months) revealed spike and slow wave activity and anticonvulsive therapy with sultiame was initiated.

At initial presentation there was pendular nystagmus, no fixation of light and right exotropia of 17°. The optic discs appeared to be slightly pale (Fig. 1b). Retinoscopy showed myopia and astigmatism of −0.5/−1/180° of the right eye and an astigmatism of −1.0/180° of the left eye. Myopia increased to −10 D in both eyes by the age of 5 1/2 years. At this age, visual acuity was 0.9 cycles per degree (cpd; Teller Acuity Cards). Derivation of monocular values was not possible as compliance was low. Under patching therapy the exotropia resolved to a parallel primary position.

The study complied with the Declaration of Helsinki and was approved by the ethics committee of the Medical Faculty, University of Regensburg. Informed consent was obtained from the subject’s guardian prior to examination.

Results

Most amplitudes were reduced below the fifth percentile of an age-matched normal control group (Fig. 2a). This was particularly true for the single-flash cone b-wave, flicker and OP responses, which were outside the normal range at all examinations, showing a maximal reduction of 55 (single flash cone b-wave both eyes), 62 (flicker right eye) and 55 (flicker left eye), and 40 μV (OPs both eyes), respectively. In the first examination OPs were indistinguishable from the average noise level. Responses of the scotopic system were well within or toward the lower limit of the normal range such as the left eye rod response at age 14 months of 142 μV, both eyes combined response a-wave of 152 and 257 μV, respectively, and left eye combined response b-wave at age 3.5 years of 294 μV. Implicit times were either within the normal range or minimally prolonged.

Fig. 2
figure 2

a Amplitudes of the five standard responses as a function of time. Only very few responses fall in the normal range indicated by the 5th and 95th percentiles (light and dark lines, respectively). Triangle Left eye, diamond right eye. b Flash VEP responses of right eye (left column) and left eye (right column) at age 1.17 (top row) and 3.5 years (bottom row). Two recordings per eye (dark trace first recording, bright trace second recording) per recording session were performed. Main features such as P2 are reproducible. OD Right eye, OS left eye

Full size image

Responses of the flash VEP were severely distorted, rendering precise implicit time measurements of the typical troughs and peaks impossible. However, gross features of responses, e.g. P2, were reproducible (Fig. 2b).

Discussion

Assuming that there is an effect of sevoflurane on the ERG, we consider these changes too small to fully account for the massive reductions in amplitudes [6]. However, an influence of the anaesthetic agent and intrasubject variability might be responsible for the decline of amplitudes observed between the last two examinations as some differences were as small as 5 μV.

Some studies have suggested a link between hyperhomocysteinaemia and microvascular damage [5]. As plasma homocysteine levels remained elevated the marked reduction in OPs may be explained by impairment of retinal microvascular circulation.

Increased homocysteine attenuates the transport of folates to photoreceptors [10], thereby decreasing the amount of folates essential for intracellular anabolic processes. Photopic ERG responses seemed to be relatively more impaired than the rod and combined responses, suggesting a preferential impact on cone photoreceptor function. However, this idea is in conflict with the hypothesis that response changes are due to alteration of microvascular circulation, whereby the same impact on both the scotopic and photopic systems would be expected.

As homocysteine is toxic to neurons of the ganglion cell layer [9] a possible ganglion cell loss might account for the severely deformed flash VEP responses. Alternatively, VEP changes might be secondary to an altered signal input into ganglion cells. Pathological responses might be partially due to a congenital malformation through elevated homocysteine [8]. However, we favour the theory of a functional rather than anatomical impact considering that both fundi displayed only minor abnormalities.

The successful patching implies that functionally relevant signal processing is in operation. Since the fovea is fully developed at an age of 5.5 years we do not expect any significant improvement in retinal function.

It is questionable whether all the observed changes are attributable to increased homocysteine levels alone as there are other diseases in which homocysteine levels are elevated to a similar degree to that in the cblG defect but with normal visual system function [2].