Food restriction alters neuronal morphology in the hypothalamic ventromedial nucleus of male rats

Abstract

Several lines of evidence have implicated the hypothalamic ventromedial nucleus (VMH) in the control of caloric homeostasis. For example, the activity of VMH neurons depends on energy availability. We tested the hypothesis that energy balance may involve the remodeling of the dendritic arbor of VMH neurons. We compared two groups of animals: one group had ad libitum access to food, and the other experienced 10-d restricted access to food. As expected, the food-deprived group lost body weight and had reduced levels of glucose, insulin, and leptin. VMH neurons were visualized after Golgi impregnation, and dendrite length was measured. Food deprivation had differential effects on VMH neurons. In particular, within the ventrolateral VMH, for neurons with long primary dendrites (LPDs) that extended in the lateral, but not medial, direction, the LPDs were 31% shorter. These same neurons exhibited a 32% reduction in the number of other dendrites without a change in soma size. In contrast, within the dorsomedial VMH, for neurons with medially, but not laterally, extended LPDs, the soma area was reduced by 28%. However, neurons in the dorsomedial VMH did not display a change in the length or number of dendrites, regardless of LPD direction. Thus, although structural changes during calorie depletion occur in both the dorsomedial and ventrolateral VMH, only the latter exhibits a remodeled dendritic arbor. These results also suggest that the direction of the LPD may be an important marker of neuronal function in the VMH.

Figure 1

Line graph illustrating the average body weights for the ad libitum control and food-deprived experimental groups. During a week of baseline measurements, all rats had free access to food and water. The food-deprived rats then had access to food only on alternate days for the next 10 d (onset indicated by arrow). The body weight of the food-deprived rats significantly decreased as the control group continued to gain weight daily. *, Significant difference between groups (Newman-Keuls, P = 0.00003 on d 17).

Figure 2

Scatter plot of the average daily food intake for the control and food-deprived rats. Food intake for the control rats did not change across these 17 d. For the food-deprived rats, food intake increased once the deprivation regimen began (indicated by arrow), reaching significance above baseline on d 14 and 16. *, Significant difference between groups (Newman-Keuls, P < 0.005 and P = 0.0001, respectively).

Figure 3

Digital photomicrograph (panel A) and the corresponding camera lucida drawing (panel B) of a representative VMH neuron. The proximal region of the LPD, with a secondary dendrite (Sec), and one of the short primary dendrites (SPDs) are indicated with arrows. Note that camera lucida drawings are based on multiple z planes, thereby depicting the extension of dendrites that are not fully visible in photomicrographs. Also note that the LPD extends beyond the view of this image. VMH neurons were photographed at ×20. Scale bar, 25 μm for both panels A and B. Scale bars are not identical because the two images are enlarged by slightly different amounts.

Figure 4

Drawing of the mediobasal hypothalamus summarizing the location of all VMH neurons in this study, including neurons from both the ad libitum (open symbols) and food-deprived (FD) (filled symbols) animals. Circles represent the neurons with LPDs extended in the lateral direction, whereas squares represent VMH neurons with LPDs extended in the medial direction. The drawing is based on Paxinos and Watson (34). 3V, Third ventricle.

Figure 5

Bar graph depicting the length of LPDs in ad libitum and food-deprived rats according to direction. In the ventrolateral VMH, LPDs extended in the lateral, but not medial, direction were significantly shortened after food deprivation. *, Significant difference between groups (P = 0.005). There was no effect of food deprivation on LPD length in the dorsomedial VMH.

Figure 7

Bar graph depicting the number of short primary and secondary dendrites (SPDs and SECs, respectively) in ad libitum and food-deprived rats, according to the direction of the LPD. Within the ventrolateral VMH, neurons with laterally, but not medially, extended LPDs exhibited a reduction in the number of other dendrites (*, P < 0.005; †, P < 0.05). There was no effect of food deprivation on dendrite number in the dorsomedial VMH (data not shown).

Figure 6

Illustration of two possible mechanisms that may underlie the observed reduction in the length of laterally extended LPDs in the ventrolateral VMH. A, Two representative neurons in the VMH in the ad libitum condition. The neuron on the right has a laterally (L) extended LPD, whereas the neuron on the left has a medially (M) extended LPD. B, Depiction of how these neurons may change morphology during food restriction, with all laterally extended dendrites becoming shorter (note that both neurons are affected). This scenario implicates a role for laterally arriving afferents. C, An alternative pattern of changed morphology, with a loss of dendrites on neurons with laterally extended LPDs dendrites (note that the neuron on the left is not affected). This scenario suggests that neurons with laterally extended LPDs are specifically involved in energy balance compared with other ventrolateral VMH neurons.

Figure 8

Bar graph depicting the length of laterally extended short primary dendrites (SPDs) and secondary dendrites (SECs) in ad libitum and food-deprived rats in the ventrolateral VMH. In either case there was no effect of food deprivation on the length of these dendrites.

Figure 9

Bar graph depicting the soma area of VMH neurons in ad libitum and food-deprived rats according to the direction of the LPD. Within the dorsomedial VMH, neurons with medially, but not laterally, extended LPDs exhibited a reduction in soma area (*, P < 0.0002). There was no effect of food deprivation on soma size in the ventrolateral VMH.