The authors regret that: 1. The published Eq. (2) is correct, but describes the release into the surrounding media and not the decreasing amount of released protein in the μ-beads. However, the latter was measured and presented in the paper. For correctness the leading term (1-) has to be eliminated. Then the correct Eq. (2) already used for fitting the release data reads as follows: cc0=6π2∑n=1∞1n2exp−DBeadn2π2ta2 2. There is a scaling error for the determined diffusion coefficients inside the μ-beads D_Bead, which occurred in the fitting process. μm²/h was wrongly used instead of μm²/s. This affects Table 2 and two statements in the main text including comparisons to diffusion coefficients in solution. The corrected D_Bead is smaller by a factor of 3600. This fact more strongly underlines the slow diffusion inside the μ-bead and the validity of our assumptions. Hence, numbers for D_Bead in Table 2 have to be corrected (see below). The main text has to be changed.The sentences (last two sentences on page 63): ‘A fit of the model to the data reveals a diffusion coefficient inside the μ-beads D_Bead of roughly 0.2 μm²/s (Table 2). This diffusion coefficient is three orders of magnitude smaller than D for free diffusion of lysozyme, which is about 100 μm²/s [57].’ should read as: ‘A fit of the model to the data reveals a diffusion coefficient inside the μ-beads D_Bead of roughly 4 × 10^–5 μm²/s (Table 2). This diffusion coefficient is six orders of magnitude smaller than D for free diffusion of lysozyme, which is about 100 μm²/s [57].’ The sentence (page 65, right column): ‘From the fits of Eq. (2) to the SDF-1 release data, SDF-1 mobility inside the μ-beads (D_Bead) was around 0.3 μm²/s and therefore in the same range as for lysozyme.’ should read as: ‘From the fits of Eq. (2) to the SDF-1 release data, SDF-1 mobility inside the μ-beads (D_Bead) was around 8 × 10^–5 μm²/s and therefore in the same range as for lysozyme.’ The authors would like to apologise for any inconvenience caused.