You’ve stumbled across an interesting phenomenon called hyperchromicity or hypochromicity (depending on whether absorbance increases or decreases). Central to understanding your question is a basic understanding of absorption phenomena.
It turns out that absorption is simply a consequence of what happens when a light ray or wave passes through something. The absorption event really is absorption of energy at a quantum level. There is a partial loss of energy form the light ray that perfectly matches the amount of energy taken up by the absorbing material. Generally, we’re talking about an electron that absorbs the energy and jumps up to an excited state. The probability that this happens is linked to how closely the energy of the particular light ray (determined by its wavelength) matches the energy gap between resting and excited states of the electron … basically, if they match, there is absorbance.
The inverse of absorption is called transmission; a medium is transparent to the extent that it allows transmission of light. The transmissivity of a medium depends on the wavelength of the radiant energy, determined by the composition of material. For example, since you’re working with DNA, you probably know already that the nucleotide bases absorb around ~280 nm. Most plastics also absorb around this wavelength, and though they may appear to be clear in our visible wavelength, they most often are opaque in the range where DNA is absorptive. (Historically, this has meant the experiment you’re talking about has been conducted in a quartz cuvette/tube, or is made of a special plastic that is “optically clear” around 280 nm.
Absorption is generally an additive property. If you take two strands of DNA and add them together at the same concentration, they are expected to give you the sum of the absorbance. Likewise, we should be able to compare a strand of DNA with the individual nucleotide-triphosphates (a mix of ATP, TTP, GTP, CTP) and come up with the same absorption spectrum if we match the concentrations right. That’s what we expect from experiments on light and transmissivity.
In practice, it is well known that a DNA or RNA double-helix absorbs less than when the two strands are separated…put another way, the transition from double-strand to single-strand causes an increase in absorbance, termed the “hypochromic effect”. What’s strange about the hypochromicity is that there seems to be some missing energy, the result is an apparent diminution of absorption when duplexed. What’s even more strange (at least it’s strange to me) is that the double-stranded to single-stranded transition takes place without any chemistry (only hydrogen-bonds between bases are broken). So what’s going on?
A truly quantitative (rigorous) understanding of hypochromism you’re likely only to find as part of a Ph.D. course in Biophysics/Physics/Chemistry. That’s because it’s a quantum phenomenon that results from the interaction of one particular electronic excited state of a given chromophore and different electronic excited states of neighboring chromophores. In fact, it’s precisely the regular structure of DNA or RNA in a double-helix that puts nucleotide bases next to one another, in close enough proximity that the excited state from one base following light absorption actually interacts electronically with the bases above and below in the helix.
As I said, why this happens is difficult to explain without using quantum mechanics (calculus), which is why you generally don’t get a detailed explanation until graduate school. I’ll try to use pictures to replace the math. The base that is absorbing a photon of light is chemically not symmetric, and so the additional energy gives rise to a dipole moment (like a bar magnet); since this dipole results from an absorption event, and a transition of an electron from low to excited state, it’s called a “transition dipole”. When the DNA is unstructured, that’s it; the neighboring bases are too far away to affect or be affected by our transition dipole, and so the absorbance you measure adds up the same way as a mixture of NTPs at the same relative concentration. In a double-helix, neighboring bases are affected by the oscillating magnetic field from the transition dipole of our absorbing nucleotide, even if they do not absorb a photon of light themselves. Because these neighboring groups are also chemically asymmetric (electronically, they are polarizable), a dipole is induced in the surrounding bases, and they take on some of the energy from the transition dipole; these induced dipoles are do not dissipate the energy of the light, because they simply reradiate (this also relates to refractive index). Rather, depending on the frequency of the induced dipoles, they may be in phase or out of phase with the phase of the exciting light; the consequence of this is “sign” is an orientational dependence.
It just so happens that in a DNA or RNA helix, the physical arrangement of bases with respect to one another sets up a situation where dipoles are aligned in a mutually repelling fashion. This makes it more difficult to create a transition dipole of the absorbing base; the result is a shorter dipole, which means less absorption or hypochromism. (Hypochromism is a consequence of planar stacking; hyperchromism would be expected for end-to-end stacking). What’s really weird about the whole thing is that there must be constant energy when intensity is integrated over all transitions and must be unaffected by interactions between states. Therefore, if the lowest-energy transition has less intensity, then there must be higher energy transitions (higher wavelengths) that must have more intensity and vice-versa. So in part, we observe a decrease in DNA absorbance of ~30% in a duplex, but only because we’ve parked our instrument at 280 nm.
Most of what I’ve discussed can be found in great detail in a series of books called “ Biophysical Chemistry” by Charles R. Cantor and Paul R. Schimmel, in particular, in volume II (ISBN 0-71671-190-7). You might also learn more from some of the better Physical Chemistry textbooks, especially those with a focus on biological sciences.
I hope this helps!
Dr. James Kranz