If you’ve ever tried to separate a real neural signal from the noise of movement artifacts or photobleaching, you’ve probably encountered the concept of an isosbestic point. An isosbestic point is a specific wavelength at which two or more chemical species, typically different conformational or binding states of a molecule, have identical molar absorptivity (or, in fluorescence terms, identical excitation efficiency). In other words, it’s a wavelength at which the fluorescence signal is insensitive to the thing you’re trying to measure. For a calcium indicator like GCaMP, the isosbestic point sits around 405–410 nm: whether calcium is bound or unbound, the dye fluoresces equally when excited with this wavelength. The signal you collect at this wavelength reflects everything except calcium, including motion, blood vessel artifacts, sensor expression levels, and fiber coupling drift, making it a perfect internal reference.

In fiber photometry, this property is exploited through a dual-excitation approach. We can simultaneously (or in rapid alternation) illuminate tissue with a calcium-sensitive wavelength (typically ~470 nm for GCaMP) and with the isosbestic wavelength (~405 nm). Both wavelengths produce emission which is collected through the same fiber, meaning both signals suffer from the same mechanical and optical noise. By dividing or regressing the 470 nm signal against the 405 nm signal, as ΔF/F or as a motion-corrected ratio, we can isolate the calcium-dependent fluorescence change with dramatically improved signal quality. Without this correction, even subtle movements of the animal can produce fluorescence transients that are virtually indistinguishable from genuine neural activity.

Figure 1. (A) Excitation spectra of calcium-bound (green) and calcium-free (blue) GCaMP. The two curves intersect at the isosbestic point (~405 nm), where fluorescence is identical regardless of calcium occupancy. A second excitation wavelength (~470 nm) is calcium-sensitive. (B) Example dual-channel photometry traces. Both channels are equally affected by a motion artifact (pink shading), allowing it to be subtracted out by regression. Only the 470 nm channel captures the true calcium event (green shading). (C) Three correction strategies depending on sensor type: optical referencing with the isosbestic wavelength (most robust), co-expression of a spectrally distinct inert fluorophore, or computational regression for sensors with no accessible isosbestic point.

The challenge arises with newer or engineered sensors that lack a well-defined isosbestic point. Many red-shifted sensors (like RCaMP or jRGECO1a), voltage indicators, and some dopamine or norepinephrine sensors (like GRAB_DA or dLight) do not have a convenient isosbestic wavelength accessible with standard hardware. In these cases, we have a few options as a workaround. One common approach is to co-express a spectrally distinct, calcium-insensitive fluorophore (such as a red-shifted tdTomato alongside a green sensor) and use its signal as the motion reference channel. Another strategy relies on careful computational regression using the raw signal itself, fitting and removing low-frequency drift and motion-correlated variance through methods like principal component analysis or robust linear regression, though these approaches require more validation and cannot fully replace an optical reference.

The absence of an isosbestic point is not a dealbreaker, but it does demand more methodological rigor. Choosing the right reference strategy depends on your sensor, your hardware, and how stringently you need to separate artifact from signal. As the sensor toolkit in systems neuroscience continues to expand at a rapid pace, developing isosbestic-independent correction methods is becoming an increasingly active area of technical innovation, and something every photometry user should have on their radar.

To learn more about the OASIS Photometry and its features, click here or contact our customer support team.