As a cloud engineer in an app development startup using Kubernetes on GKE, you're faced with a challenge. Several applications are deployed on separate VPC-native Google Kubernetes Engine clusters within the same subnet, which has no more available IPs. What solution ensures these clusters can expand their nodes when necessary?