The Internet of Vehicles (IoV) can offer safe and comfortable driving experience, by the enhanced advantages of space-air-ground integrated networks (SAGINs), i.e., global seamless access, wide-area coverage and flexible traffic scheduling. However, due to the huge popular traffic volume, the limited cache/power resources and the heterogeneous network infrastructures, the burden of backhaul link will be seriously enlarged, degrading the energy efficient of the IoV in SAGIN. In this paper, to implement the popular content severing multiple vehicle users (VUs), we consider a Cache-enabled Satellite-UAV-Vehicle Integrated Network (CSUVIN), where geosynchronous earth orbit (GEO) satellite is regard as a cloud server, unmanned aerial vehicles (UAVs) are deployed as edge caching servers. Then, we propose an energy-aware coded caching strategy employed in our system model to provide more multicast opportunities, and to reduce the backhaul transmission volume, considering the effects of file popularity, cache size, request frequency, and mobility in different road sections (RSs). Furthermore, we derive the closed-form expressions of total energy consumption both in single-RS and multi-RSs scenarios with asynchronous and synchronous services schemes, respectively. An optimization problem is formulated to minimize the total energy consumption, and the optimal content placement matrix, power allocation vector and coverage deployment vector are obtained by well-designed algorithms. We finally show, numerically, our coded caching strategy can greatly improve energy efficient performance in CSUVINs, compared with other benchmarked caching schemes under the heterogeneous network conditions.