I.The cause of the quantum entanglement is that the wave-function (w.f.) of the involved particles doesn't have the form of the w.f. of independent particles.
So, let's begin by defining what is the w.f. of independent particles. The w.f. of a single particle, let's call it A, can look like
$$|A> = \frac {\lvert x_A \rangle + \lvert y_A \rangle.} {\sqrt(2)}$$
where x and y can be the possible polarizations of the particle if it has spin, or it is a photon.
Similarly, for the particle B
$$|B> = \frac {\lvert x_B \rangle + \lvert y_B \rangle.} {\sqrt(2)}$$
So, the w.f. of these two particle should be
$$|A>|B> = \frac {(\lvert x_A \rangle + \lvert y_A \rangle)(\lvert x_B \rangle + \lvert y_B \rangle)} {\sqrt(2)}$$
Opening parentheses,
$$ |A>|B> = \frac {\lvert x_A \rangle \lvert x_B \rangle + \lvert x_A \rangle \lvert y_B \rangle + \lvert y_A \rangle \lvert x_B \rangle + \lvert y_A \rangle \lvert y_B \rangle} {2}$$
So, this is the w.f. of the two particles if they are independent i.e. what happens with one particle has no connection with what happens with the other. We use to say that this form is factorizable, i.e. can be represented as a product in which each factor refers to a single particle.
In an entangled pair of particles one of the products of states is missing, e.g.
$$|A, B> = \frac {\lvert x_A \rangle \lvert x_B \rangle + \lvert y_A \rangle \lvert x_B \rangle + \lvert y_A \rangle \lvert y_B \rangle} {\sqrt(3)}$$
We cannot factorize this state.
This is what changes.
II. How the information is exchanged between A and B, we don't know, we only know the result of this supposed exchange of information. The result, see the non-factorizable w.f. is as follows:
If we measure the polarization of the particle A and find it along x, then a measurement of the polarization of B can be found only y. But if we find for A, polarization y, then for B we can find either x, or y. Similar things we can say about measuring the polarization of B and what happens with A.
III. One can entangle not only 10 particles, much more. Though, it will be difficult to work with that wave-function. People study much simpler entanglements, i.e. of 2, 3, 4, particles.
There is a theorem of no-communication that says
NO ACTION DONE ON ONE OF THE PARTICLES HAS AN EFFECT ON THE OTHER PARTICLES.
IV. Operations done on particles can be reversible, or non-reversible.
An example of irreversible operation is to measure the polarization of one particle as in our example in part I. Yes, such an operation breaks the entanglement for all the particles entangled with the measured particle. On the other hand, a reversible operation, e.g. passing one particle through a classical field, has no implication on the entanglement. he entanglement remains entire.