Fig 1.
Brainstem cutting planes and transverse sections.
(A) Brainstem sagittal view demonstrating cutting planes. The blocks containing the rostral interstitial nucleus of the medial longitudinal fascicle (RIMLF), the oculomotor nucleus (nIII), the paramedian pontine reticular formation (PPRF) including the excitatory (EBN) and inhibitory burst neurons (IBN), the nucleus raphe interpositus (RIP) containing omnipause neurons (OPN), and the abducens nucleus (nVI), were cut in series of 10μm and 5μm thick sections. (B) Caudal view of the block containing the RIMLF. (C) Caudal view of the block containing the PPRF and OPN region. Scale bar B,C = 1cm. INC, interstitial nucleus of Cajal; IO, inferior olive; MB, mammillary body; LGN, lateral geniculate nucleus; MCP, medial cerebellar peduncle; MGN, medial geniculate nucleus; MT, mammillothalamic tract; nIV, trochlear nucleus; NVI, abducens nerve; PC, posterior commissure; RN, red nucleus; TR, tractus retroflexus; PC, posterior commissure; PUL, pulvinar; SC, superior colliculus; SCP, superior cerebellar peduncle; SN, substantia nigra.
Fig 2.
Histologically normal omnipause neurons.
Omnipause neurons in the nucleus raphe interpositus (RIP) appear histologically normal (arrows). (A) Photograph of a transverse section of the pons at the level of RIP. The inset indicates the area shown in A. (B) LFB-PAS staining showed no evidence of demyelination. (C) CR 3/43 staining showed no abnormal microglial activation. (D) GFAP staining showed no reactive gliosis. The asterisk labels the same blood vessel in neighboring sections for orientation. LFB-PAS, Luxol fast blue periodic acid-Schiff; GFAP, glial fibrillary acidic protein. Scale bar A = 500μm, inset = 5mm; B-D = 200μm.
Fig 3.
Normal abducens nucleus perineuronal net immunostaining.
Detection of perineuronal nets by aggrecan (ACAN) immunolabeling in the abducens nucleus (nVI) appeared similar in control subject and patient specimens. Low power (A and B) and high power (C and D) views of parvalbumin-positive (PAV) abducens neurons (brown) ensheathed by aggrecan (ACAN)-immunoreactive perineuronal nets (black, arrows). In the tissue from the control (A and C) and from the patient (B and D), ACAN-stained perineuronal nets (arrows) appeared in unaltered manner. Low power (E and F) and high power (G and H) photographs of the inhibitory burst neurons (IBN) of the control (E and G) and patient (F and H) displaying similar perineuronal nets around PAV-positive IBN. Scale bars A,B,E,F = 100μm; C,D,G,H = 30μm.
Fig 4.
Absent or fragmented omnipause neuron perineuronal net immunostaining.
Perineuronal nets around omnipause neurons (OPN) were absent or abnormal by all immunostaining methods, compared to a normal control subject. The left column demonstrates a control subject, whose parvalbumin (PAV)-positive OPN appear normal. Low- and high-power views of the control show normal-appearing perineuronal nets staining with HPLN1 (A, C, arrows) and aggrecan (ACAN) antibodies (E, G). The right column demonstrates OPN stained for PAV (B, D, F, H, arrows). Their perineuronal nets are either absent or fragmented when stained for HPLN1 (B, D, arrows) or ACAN (F, H, arrows). High-power views of PAV-positive OPN (brown) contacted by synaptophysin-positive terminals (black) in the control (I) and patient (J) appeared similar. Scale bars A,B,E,F = 100μm; C,D,G,H,I,J = 30μm. Dashed line (A, B, E, F) designates the midline.
Fig 5.
Absent or fragmented omnipause neuron perineuronal net triple immunofluorescence staining.
Triple immunofluorescence staining for different components of perineuronal nets, revealed by a confocal laser scanning microscope. In the control case, omnipause neurons (OPN) are ensheathed by prominent perineuronal nets showing the same appearance with antibodies against the link protein (HPLN1), chondroitin sulfate proteoglycan (CSPG) and aggrecan (ACAN) (A, D, G, arrow). In the saccadic palsy patient, the neurons of the superior olive (SO) from the same sections as OPN are ensheathed by prominent perineuronal nets revealed by immunostaining of HPLN1, CSPG, and ACAN (C, F, I, arrows). However, around OPN (asterisk) in the patient, only HPLN1-based perineuronal nets can be detected, which appear fragmented (B, arrow). CSPG- and ACAN-immunostaining does not reveal perineuronal nets, but only few fragments along a few dendrites (E, H, arrow). Scale bars A,D,G = 20μm; B,C,E,F,H,I = 200μm.
Fig 6.
Fragmented burst neuron perineuronal net immunostaining.
Parvalbumin-positive (PAV) putative burst neurons in the rostral interstitial nucleus of the medial longitudinal fasciculus (RIMLF) and in the paramedian pontine reticular formation (PPRF) were present in both the control and patient. Perineuronal nets appeared normal around burst neurons (EBN) in the PPRF (A, C, arrows) and RIMLF (F, H, arrows) in the control, here immunostained for aggrecan (ACAN). However, the patient’s burst neurons in the PPRF were ensheathed only by fragments of ACAN-based perineuronal nets (B, D, short arrow). Few PAV-positive neurons were found with preserved perineuronal nets (B, E, long arrow). Similarly, the PAV-positive neurons (brown) in the RIMLF of the control are enwrapped by distinct perineuronal nets (F, H, arrows), whereas only fragments are found in the RIMLF of the patient (G, short arrows). Most PAV-positive neurons are attached to few ACAN-positive fragments (I, short arrow). Scale bars: A,B,F,G = 200μm; C,D,E,H,I = 30μm.