My first impressions of UV light and its mechanics diverge a little -- as far as I'm aware, UV light does not have mutation-promoting effects, but outright destructive effects on DNA and various molecules, especially long polymers. The various components of UV cause sun-bleaching and direct-light wear. It might be more difficult expressing genetic damage without an intracellular metabolism, but I propose UV light:
- Wanes with distance from the surface of the water (if such an option is enabled). This lowers the probability of damage occurring.
- Does not promote mutation nor does it outright kill every cell- rather, it has a chance to "sever" inheritance, preventing a cell from making an adhesin connection when it splits or erasing its descendant modes, forcing one or both of its successor cells to be the same mode as their parent. It means cells can still survive in the presence of UV even without protection, but will be forced to remain fairly simple if not uni-cellular.
- As an abstraction of the cell's effort to repair genetic damage, direct UV exposure consumes additional energy. Alternatively, UV light exposure slows down the spread of signals, nutrients, nitrogen, etc. across adhesin bonds.
- Further, melanin cells should allow light into adjacent or connected cells whilst blocking UV, to advantage melanin-protected photocytes while also simulating a form of foliage competition.
- As a bonus, UV light could bleach the cell's actual colors over time, gradually inching them towards a white color.
Oh, and one mode suggestion for the road: (Hemo)Lymph node.
A glandular cell that secretes resource-transporting hemolymph in its front-facing direction. The hemolymph travels along adjacent cells –- not through adhesin connections, but parallel to them, as if adhering to the cell walls themselves –- Every lymph node produces a set "volume" of hemolymph, advantaging cellular arrangements with closer adhesin connections where the hemolymph may flow- but not where cells outright touch; the distance between adhesin connections is what determines the "range" of the hemolymph flow. The lymph node "pushes" nutrients and signals out to all cells it touches with its hemolymph, including other lymph nodes, though it may be set to only produce hemolymph which conducts one or the other- giving potential for separated vascular/neural systems. (A bidirectional flow of resources would probably be less interesting than a one-way flow, as determined by the lymph node and its connections.)
The inspiration for the lymph node, or equivalent, is that of a primitive internal body cavity and a shared vascular system, as is present in all organisms more complex than multi-cellular colonies. It'd be a fairly important distinction, especially when melanin and other dermal cells are concerned, but it'd shine most in the case of additional resources being implemented (cellular waste, toxins, cellular products, and others.)