4:4:4 refers to bit depth captured by the camera.
4:4:4 has noting to do with bit depth, I think you're confusing it with something else.
When you see references to numbers like 4:4:4, 4:2:2, 4:1:1, etc these are basically ratios indicating the number of samples stored per color channel in each frame of your video.
In video color and brightness information is encoded in three channels - generally denoted as "YUV" - with "Y" representing the luminance (brightness) and "U" and "V" representing chrominance (color) information.
The process of video compression involves discarding information from the signal that is not likely to be visible to the viewer. The human eye is most sensitive to changes in luminance and much less sensitive to changes in chrominance. Thus, the first step in image compression is generally to reduce the amount of data stored for the chrominance channels because it's not likely to be missed by the viewer.
So with 4:4:4 color sampling each line of four pixels in the image has four pixels in each of the three channels - this is 'uncompressed' as far as the channels are concerned (other types of compression may still be applied to the image later in the pipeline).
With 4:2:2 each four pixel segment stores four pixels in the Y channel and only two pixels each for the U and V channels. This reduces the total data you need to store by 33%.
As you've probably guessed, 4:1:1 has only one pixel in the U and V channels for every four in the Y channel. Now we've cut the total data in half.
The one odd (and slightly confusing) exception to this is 4:2:0, which is a common color sampling format for many consumer/prosumer video formats. By the normal logic of the other ratios you would expect this to indicate four pixels of Y, two of U, and no V - but that would result in an image that looked completely wrong. 4:2:0 is actually a variation on 4:1:1 - it just alternates the chrominance samples across two adjacent lines. So the first four pixels of your first line have four pixels in Y, two in U, and none in V; the first four pixels of your second line have four Y, no U, and two V. This results in exactly the same number of total samples as 4:1:1 and has the equivalent 50% data reduction.
This reduction is just the first step in the compression pipeline. By reducing the initial amount of data you allow the next step in the process to work more efficiently, which generally results in fewer visible errors in the final compressed video. Thus for cameras that compress to a low-to-moderate bit rate it's common to use at most 4:2:2, and more commonly 4:1:1 or 4:2:0, in order to improve the overall visual appearance of the compressed video.
Compression is always a trade-off between file size and picture quality. Reducing the color channel information generally works well because our eyes aren't particularly sensitive to it - but there are situations where it can introduce visible errors. For instance, 4:2:0 works great for most natural scenes, but shoot in a nightclub lit primarily by strongly saturated blue or red lights and you'll get an image that looks soft and blocky because there's not enough pixels in the color channels to accurately represent the detail in the scene - so for a specific situation like that you may need to select a camera that records at least 4:2:2. Color keying - green or blue screen - relies on the color information to create a clean separation of the subject from the foreground; the more information you have in the color signal the better the end result. So for projects that rely heavily on chroma keying it may be worth selecting a camera setup that records a full 4:4:4 signal - often this will require a tethered computer or recorder because it's uncommon for a camera to do it on board.
As far as it relates to the original topic of lighting... in general, it doesn't really. Unless, of course, you're lighting a scene primarily with highly saturated red or blue lights.
